riparian management classification for canterbury streams · riparian management classification for...

Riparian management classification for Canterbury streams

NIWA Client Report: HAM2003-064 June 2003 NIWA Project: ECN03202

All rights reserved. This publication may not be reproduced or copied in any form without the permission of the client. Such permission is to be given only in accordance with the terms of the client's contract with NIWA. This copyright extends to all forms of copying and any storage of material in any kind of information retrieval system.

Riparian management classification for Canterbury streams John M. Quinn

Prepared for

Environment Canterbury

NIWA Client Report: HAM2003-064 June 2003 NIWA Project: ECN03202 National Institute of Water & Atmospheric Research Ltd Gate 10, Silverdale Road, Hamilton P O Box 11115, Hamilton, New Zealand Phone +64-7-856 7026, Fax +64-7-856 0151 www.niwa.co.nz

Contents

Executive Summary iv 1. Introduction 1

2. Methods 3 2.1 Survey methods 3 2.2 Riparian function assessment protocols 5 2.2.1 Bank stabilisation 5 2.2.2 Filtering contaminants from overland flow 6 2.2.3 Nutrient uptake by riparian plants 7 2.2.4 Denitrification 8 2.2.5 Shading for instream temperature control 8 2.2.6 Shading for instream plant control 9 2.2.7 Input of large wood and leaf litter 10 2.2.8 Enhancing instream fish habitat and fish spawning areas 10 2.2.9 Controlling downstream flooding 11 2.2.10 Human recreation 11

3. Results 12 3.1 General characteristics of the Canterbury study sites 12 3.2 Assessment of riparian functions 16 3.3 Predicting riparian function activity 17 3.3.1 Predicting current riparian functions 17 3.3.2 Predicting potential riparian functions 20 3.3.3 Predicting individual riparian functions 25 3.3.4 Geomorphic approach to riparian management classification 27 3.3.4.1 Riparian/stream geomorphic classes 30 3.3.4.2 Other potential classification attributes 34 3.4 Application of riparian management classification in river management 36

5. Acknowledgements 37

6. References 38

7. Appendix 1: 44

8. Appendix 2: 53 ________________________________________________________________________ Reviewed by: Approved for release by:

Stephanie Parkyn Kevin Collier

Formatting checked

………………………

Riparian management classification for Canterbury streams iv

D R A F T

03

Executive Summary

I. Riparian zone management provides opportunities to mitigate damage to the ecological health and human uses of streams, rivers and downstream aquatic ecosystems caused by intensive production land uses. However, the functions of riparian zones vary spatially within catchments, so their management requires a framework that matches actions with the functions occurring at a site, and with river management goals. This study aimed to develop ways of classifying riparian areas within relatively large catchments according to their functional roles in improving stream habitat, controlling contaminant inputs and enhancing aesthetics, biodiversity, and recreation. When linked with information about river goals, such classifications should be useful tools for planning and prioritising riparian management actions.

II. Rapid assessments were carried out over usually 100 m long reaches at 313 sites to evaluate site characteristics (physical, vegetation and management practices) and use a protocol developed as part of the study to rate the activity of 12 riparian functions (i.e., streambank stabilisation; filtering contaminants in overland flow; nutrient uptake from shallow groundwater; denitrification of shallow groundwater; shade for instream temperature and nuisance plant control; input of wood and leaf litter to the stream; fish habitat enhancement; control of downstream flooding; and enhancing recreation and stream aesthetics). Both current and potential (under best practicable riparian management) ratings were made. These sites covered a representative range of the areas of agriculture and forestry in Canterbury, during late spring 2000 and 2002 and early summer of 2001 and 2003. Sites ranged from small, 1st order, headwater streams to large, 7th order, braided rivers. Additional catchment and site data were obtained from the River Environment Classification (REC) and Land Environments of New Zealand (LENZ) databases.

III. The survey found a low level of stream fencing (only 12% fenced on both sides) and generally low ratings for most riparian functions. Grass was the most common dominant riparian vegetation type (48% of sites), followed by willows (26%), low shrubs (9%), and native trees (8%).

IV. The current activity of twelve riparian functions varied widely between the sites but was typically rated as very low for input of woody debris, enhancement of fish habitat and recreation, and denitrification of groundwater, and “low-moderate” for bank stabilisation, control of downstream flooding, leaf litter input, shade control of stream temperature and instream plant growth, and aesthetics. Applying best practicable riparian management was judged to be capable of improving most riparian functions substantially. The biggest improvements were predicted for the shading functions and the least for denitrification.

Riparian management classification for Canterbury streams v

D R A F T

03

V. The sites were classified according to both their current and potential (RMC-P) riparian functions (e.g., streambank stabilisation, shading for temperature control, etc.) into current and potential Riparian Management Classes (RMC-C and RMC-P). Discriminant function models were developed that can be used to classify new sites (and hence infer their likely riparian function ratings), based on either GIS information alone or GIS and local information. These models provide a means for non-experts to rate riparian functions and to map the predicted riparian classes throughout catchments. Maps of predicted potential classifications to a coarse (3 class) and finer (12 class) level are provided as GIS layers on a CD. Maps showing the results at a scale that is readable in report form are presented as an example.

VI. Models were also developed to predict the activity ratings as assessed following a detailed protocol of the 12 individual riparian functions (e.g., bank stabilisation or provision of shade to control instream temperature) assessed in the field surveys. These will be useful when management is focused on particular instream issues related to a reduced number of riparian functions. For example, if high water temperatures in summer were identified as the critical factor limiting stream health, then the model that predicts where riparian management has the potential to provide shade to control temperature would be of particular relevance. GIS layers for the function “shade for stream temperature control” are provided on CD, and a map showing the results at a scale that is readable in report form are presented as an example.

VII. The statistical approach to site classification described above indicated that the key factors influencing potential riparian functions that could enhance stream habitat and water quality in the Canterbury area are related to channel width, adjacent land slope and whether the stream is ephemeral or perennial. For example, these factors control the ability of riparian vegetation to influence instream habitat (width, permanence of flow), control downstream flooding (adjacent land slope), and the importance of overland flow in transport of contaminants to the stream (adjacent land slope). This provided the basis for development of a geomorphic riparian management classification (RMC-G). Sites were classified into 12 classes based on combinations of valley-form (plain, U- and V-shaped) and 4 channel width classes in relation to the ability of different vegetation types to shade the channel. Models were also developed to predict the geomorphic classification of sites from GIS data.

VIII. A flow-chart is provided showing how the Riparian Management Classification (RMC) and microhabitat-based native plant recommendations can contribute to improved river management planning when combined with information on stream and land management goals at various spatial scales.

Riparian management classification for Canterbury streams 1

1. Introduction

Riparian zones are the three-dimensional zones of direct interaction between terrestrial and aquatic ecosystems (Gregory et al. 1991). A variety of biophysical functions of riparian areas can be managed to enhance stream habitat and water quality, including: stream bank stabilisation, filtering overland flow, shading for stream temperature and nuisance plant control, woody debris inputs, spawning habitat and cover for some fish species, and denitrification and nutrient uptake from shallow groundwater (e.g., Collier et al. 1995). Riparian areas are also heavily used for recreation, and play an important role in stream aesthetics. Their location, at the land-water interface, and the biophysical processes that occur within riparian zones, enable the management of the relatively small amount of land in riparian areas to have a disproportionately large role in controlling the effects of broader catchment activities on streams and downstream aquatic ecosystems (Collier et al. 1995, MFE 2001).

Riparian management is recognised as an important aspect of water management in the policy statements of all New Zealand’s regional councils (Boothroyd and Langer, 1999). Most proposed regional plans include a range of methods for promoting riparian management, including funding part of the costs of riparian management activities undertaken by farmers. For example, Environment Waikato has recently funded a $10 M Clean Streams Fund that focuses on riparian management to improve stream health (http://www.ew.govt.nz/ourenvironment/water/cleanstreams.htm). However, the biophysical roles and human uses of riparian margins change from headwater streams to lowland floodplain rivers, and planners need a framework that accounts for these variations so that management actions at a site are matched with riparian functions.

Quinn (1999) developed a riparian zone management classification (RMC) for the Piako and Waihou River catchments in the Waikato region that provides such a framework. This recognised 10 riparian classes based on the physical characteristics of 30 sites. The relative importance of the main riparian functions in each area was assessed in each of these 10 classification groups, and this ranking was used to recommend different riparian management options for each class. This classification provided the basis for an analysis of the costs and benefits of applying “first step” and “best practicable” riparian management options to the pasture streams in the Piako Catchment (Brown and Mackay 2000, Quinn et al. 2001). This approach was developed further in a preliminary RMC of three areas of Canterbury: Banks Peninsula; Canterbury plains catchments near Christchurch (Cam, Eyre, Halswell and Cust); and the foothill catchments of the Ashley, Hurunui and Waipara (Quinn and Suren, 2001; Quinn et al. 2001).


In this report we develop riparian management classification further by: extending the RMC approaches used in the pilot study to cover most of the area of Canterbury that is used for pastoral agriculture and forestry.


2. Methods

2.1 Survey methods

Three hundred and thirteen sites that cover the range of riparian conditions present within areas Canterbury used for pastoral/forestry land use were surveyed to provide information for developing a Riparian Management Classification (RMC, see APPENDIX 1 for site location details). Sites that could be readily accessed from roads were selected to give a good coverage of the different areas within the region. Site characteristics that affect key riparian functions and human uses were assessed over a reach at each site (typically 100 m long), photographs were taken and representative site cross-sections were sketched. Data were also collected on the stream/riparian physical attributes at three different spatial scales: catchment scale, valley scale, and reach scale (Table 1). Some of this information (e.g., valley slope, stream channel slope, % banks undercut, local land use, streambed substrate material and water width) was assessed in the field. Other data were obtained from the River Environment Classification (REC) database that NIWA has developed (Snelder et al. 1999) (e.g., catchment area, stream source of flow, geology, dominant catchment landuse, channel slope, Strahler stream order, and stream morphology class) or the Environmental Domains database (land-slope, particle size class, drainage class) and converted into numeric indices for use in statistical analyses (Table 1).

On-site assessments were made of the current functions of riparian vegetation in terms of streambank stability, denitrification of groundwater inflows, shading of the channels for temperature and instream plant control, wood and leaf litter input, enhancement of fish spawning and general fish habitat, downstream flood control, recreational use and aesthetics. The potential roles of these functions, if best practicable riparian management was applied, were also assessed. Best practicable riparian management was assumed to involve fencing out stock from the stream/riparian area and managing the area for the development of long grasses, shrubs and/or trees within this protected area as appropriate for the location (e.g., in the MacKenzie Basin we expected tussock grasses and matagouri to dominate protected riparian areas rather than large trees). These current and potential riparian functions and human uses were ranked as: 0 (absent), 1 (very low activity), 2 (low-moderate activity), 3 (moderate activity), 4 (high activity) or 5 (very high activity).

These riparian function assessments formed the basis for the riparian management classification (RMC) of sites into classes with similarities in current and potential riparian functions, and hence riparian management options. The site biophysical data provide the basis for predictive modelling of RMC classes within catchments and the entire Canterbury Region. Whether or not the function/use is important for the stream depends on the goals of watershed management, and is a separate issue to whether the function/use is active at the site.


Table 1: Details of physical attributes that describe the stream at the catchment scale, valley segment scale, and stream reach scale. # = see Snelder et al. (1999).

Spatial scale

Physical attribute

Explanatory notes

Catchment Source of flow (SOF) index Lake and lowland = 1; Hill = 2; Mountain and glacial Mountain = 3

Dominant catchment baserock geology index

Soft sedimentary =1; Alluvium & sand = 2; Miscellaneous = 3; Volcanic basic = 4; Hard sedimentary = 5

Catchment spatial average slope (°) Accumulated flow (m3s-1) Calculated upstream annual rainfall - evaporation Catchment area (km2) Catchment land cover index Bare = 1; Urban = 2; Pasture = 3; Tussock = 4; Exotic

forest = 5; Scrub = 6; Indigenous forest = 7 Valley Segment

Riparian land use Cattle, Conservation, Crop, Dairy, Forestry, Horticulture, Sheep, Urban

Channel shape category 1 = Channelised; 2 = Straight; 3 = Meandering; 4 = Sinuous

Valley bottom width category 1 = < 20 m; 2 = 20 – 50 m; 3 = 50 – 200 m; 4 = 200 – 100 m; 5 = >1000 m; 6 = “plains”

REC channel slope (cm/m) Inter-node difference in elevation/reach length REC segment mean air temperature

(°C) Predicted local mean air temperature

REC segment annual rainfall (mm) Predicted local mean rainfall temperature REC average land slope of segment’s

local catchment segment (m/m) Derived from REC digital elevation model for the land draining directly to the local stream segment

REC reach elevation (m) Above sea level Domain land drainage class 1 = Very poor; 2 = Poor; 3 = Impeded; 4 = Moderate; 5 =

Good Domain soil age class 1 = Recent, 2 = Older Domain acid soluble P class 1 = Very low, 2 = Low, 3 = Moderate, 4 = High, 5 = Very

high Domain exchangeable Calcium class 1 = Low, 2 = Moderate, 3 = High, 4 = Very high Domain induration (hardening) 1 = Non-indurated; 2 = Very weakly; 3 = Weakly; 4 =

Strongly; 5 = Very strongly indurated Reach Water width Estimate of the average wetted stream width at low flow Non-vegetated width Estimate of channel width lacking terrestrial vegetation Bankfull width Total width at bankfull discharge Wet/dry index 0 = Dry channel, 1 = Water present in channel Channel slope index 1 = < 0.2o; 2 = 0.2o – 0.5o; 3 = 0.5o – 1.0o; 4 = 1.0o – 2.0o;

5 = 2.0- 4.0o; 6 = > 4 o Local land-slope index 1 = <2o; 2 = 2.0o – 5.0o; 3 = 5.0o – 15.0o; 4 = 15.0o –

25.0o; 5 = 25 – 35o; 6 = >35° Local land slope length index 1 = plains and ≤10 m; 2 = >10 – 50 m; 3 = >50 – 200 m; 4

= >500 m Substrate composition Bedrock, boulder, cobble, gravel, sand, silt, clay Shade ratio Bank + vegetation height/ channel width


Spatial scale

Physical attribute

Explanatory notes

Bank height (m) Estimate of average bank height Periphyton categories: 0 = None; 1 = sSippery; 2 = Obvious; 3 = Abundant; 4 =

Excessive (> 80% FGA) Macrophyte species and % cover Species present, % total bed covers. Bryophyte cover

noted separately. Woody debris index 0 = Absent; 1 = Sparse; 2 = Common; 3 = Abundant Stock access index 0 = No access; 1 = One bank; 2 = Access to both banks Stock bank damage index 0 = None; 1 = Minor; 2 = Moderate; 3 = Extensive Streambank stability Assessment of the % of banks stable undercut or

slumped Riparian veg. & bank cover List of dominant riparian vegetation Dominant riparian vegetation index 0 = Bare ground, 1 = Grass, 2 = Wetland, 3 = Low shrub,

4 = High shrub, 5 = Deciduous, 6 = Willows, 7 = Coniferous, 8 = Eucalyptus; 9 = Native

Riparian wetland index 0 = absent, 1 = present Stock fencing stream index None = 0; One side fenced = 1; Both sides fenced = 2 Stock damage classes 0 = None; 1 = Minor; 2 = Moderate; 3 = Extensive Riparian fencing % of each bank fenced and bank to fence distance Fencing type index 0 = None; 1 = Electric 1 wire; 2 = Electric 2 wire; 3 = Post

& batten, or 5-7 Wire electric, or Deer fence

2.2 Riparian function assessment protocols

This section summarises the rationale for assessing each riparian function. Our assessments did not include the riparian zone functions of enhancing terrestrial biodiversity, providing wildlife corridors, and habitat and landscape connectivity, which were beyond the scope of the available resources.

2.2.1 Bank stabilisation

The role of riparian vegetation in stabilising banks depends on the ability of vegetation to: (1) reinforce bank strength through root network strengthening (Rutherfurd et al. 1999; Lyons et al. 2000), (2) provide a well-developed turf or a dense root system that protects against surface soil erosion (Murgatroyd and Ternan, 1983; Dunaway et al. 1994), (3) pump out water from the soil, and provide macropores for drainage, lowering erosion potential owing to bank sloughing and slumping (Thorne, 1990), and/or (4) buttress the toe of the streambank protecting it from shear failure (Thorne 1990). Key factors influencing these stabilising functions are: the height of the streambanks relative to the depth of root penetration, bank angles, the erosive power of the stream under high flows (including local effects such as whether the reach is straight or meandering with many erosion-prone


bends), and whether the banks are protected by other features (e.g., boulders, bedrock or large woody debris).

Grasses, herbs and forbs are expected to provide good stabilisation of small banks (< 0.5m) and those with low angles (< 45°), whereas shrubs and trees give better protection for higher and steeper banks (Burckhardt and Todd, 1998; Abernathy and Rutherfurd, 1999). The following notes provide guidance for assessing the height of streambank that can be effectively strengthened by vegetation roots (Abernathy and Rutherfurd, 1999).

• Groundcover (typically up to 1 m high including prostrate shrubs, grasses, sedges and forbs) provide reinforcement of banks to a depth < 0.3 m.

• Understorey trees (typically 1-5 m high) have roots down to about 1 m and extend laterally to about the dripline.

• Overstorey species have a central rootball or rootplate of dense roots that can usually be considered as half a sphere that has a diameter 5 times the diameter of the trunk. Root density declines rapidly beyond the root ball and for reinforcement purposes there are usually few roots beyond the canopy dripline or below about 2 m under bank surface. Watson et al. (1999) report maximum root depths of 1.8 - 3.1 m for 8 to 25 year old Pinus radiata and 1.3 - 1.6 m for 6 to 32 year old kanuka. The root stabilization function will be greatest where the bank height is less than the depth of root penetration.

2.2.2 Filtering contaminants from overland flow

To be effective at filtering of contaminants from overland flow, the riparian zone needs to: (1) slow the flow of surface runoff, enhancing settling of particulates; and/or (2) increase infiltration into the soil, enhancing filtration of particulates (Phillips, 1989a,b; Smith, 1989; Cooper et al. 1995; Williamson et al. 1996; Lowrance et al. 1997). These filtering and settling functions are enhanced by the zone having flat topography, dense ground cover of grassy vegetation or litter under riparian forest that increase surface roughness, and soil characteristics that increase hydraulic conductivity (low compaction, high sand content, abundant macropores). Obviously, the zone must receive surface runoff from the adjacent landscape for this filtering role to operate. The function will be compromised if the surface runoff is channelised, so that runoff passes rapidly through the riparian area with little time for settling of particulates or infiltration into riparian soils. The likelihood of surface runoff occurring increases with rainfall intensity, slope length, slope angle and convergence of flows, and decreases with infiltration rate. Animal trampling typically reduces infiltration


rate (Nguyen et al. 1998) and excluding stock from the riparian reverses this effect (Cooper et al. 1995). The quantity of sediment carried in surface runoff increases with the clay content of the soil. Guidelines are available to predict the optimal width of grass strip to filter suspended sediment from surface runoff in relation to slope length, slope angle, drainage and clay content (Collier et al. 1995).

2.2.3 Nutrient uptake by riparian plants

Nutrient uptake by riparian plants is an important function where infiltration surface runoff or shallow groundwater passes through the root zone before entering the stream (Fig. 1). In contrast, the function is unimportant where groundwater bypasses the root zone of riparian plants. This may occur in deeply incised streams, where tile drains deliver most of the shallow groundwater directly to the stream, or where deep groundwater emerges in the streambed as springs (Hill, 1996; Prosser et al. 1999).

Riparian vegetation type influences this function via vegetation rooting depth in relation to bank height and groundwater flows – larger trees and shrubs have deeper roots that can intercept deeper groundwater. Large plants also have a greater biomass and hence generally store more nutrient in plant tissue than small plants. Harvesting of these plants (e.g., by timber harvest or controlled animal grazing and subsequent removal of the animals) contributes to long-term removal of these stored nutrients from the riparian area. Plants nearest the stream are most likely to interact with groundwater, but nutrient uptake is expected to increase with the width of the zone of deep-rooting riparian plants.

The transpiration of riparian vegetation can also pump water from the riparian soils, leading to hydraulic gradients that draw river water into the riparian area where it is exposed to nutrient uptake and removal processes.


Figure 1: Schematic showing the influence on channel shape on the interaction between shallow groundwater and the root zone of riparian vegetation.

2.2.4 Denitrification

Denitrification is a process by which bacteria reduce nitrate to nitrous oxide and N2 gases that are lost to the atmosphere, providing permanent N removal from the water (Hill, 1996; Willems et al. 1997). The process requires nitrate N, low oxygen conditions provided by waterlogged soils or hot spots of buried organic matter, and an available carbon source to drive the process (Knowles, 1982). Denitrification is most important in riparian areas where shallow groundwater passes through wetlands before emerging in the stream (Cooper, 1990; Prosser et al. 1999). Riparian plants enhance the process as their roots increase the supply of carbon at depth within the streamside soils.

2.2.5 Shading for instream temperature control

Cool groundwater entering shallow streams heats quickly under direct solar radiation in unshaded conditions (Quinn et al. 1992; Rutherford et al. 1997; Rutherford et al. 1999). The rate of heating decreases with stream depth, as the mass of water absorbing the incident radiation increases, and with shading vegetation, that absorbs and reflects much of the incident radiation. The ability of riparian vegetation to shade the channel decreases with stream width and the height of the vegetation (Davies-Colley and Quinn, 1998). Mature trees produce a closed canopy over channels narrower than about 6 m but the shade gap

Incised channelRoot uptake unimportant

Root uptake important


between the trees on either bank increases above this width (Davies-Colley and Quinn, 1998). Tussock grasses, sedges and flaxes only provide effective shade in very narrow channels (i.e., < c. 2m). Streams with poorly conductive beds (e.g., clay or bedrock) are expected to heat more rapidly than equivalently shaded streams with conductive beds (e.g., gravels), due to less conductive heat loss to the ground and less heat exchange with groundwater. Streambanks and hills can also provide topographic shade, independent of riparian vegetation, and are particularly important in incised streams (Rutherford et al. 1999).

2.2.6 Shading for instream plant control

Riparian shade can control stream lighting and thus control instream plant growth below nuisance levels, whilst maintaining the biodiversity benefits and desirable functions that plants provide (Biggs, 2000). Shading of 60-80% is expected to prevent proliferation of filamentous green algae (Quinn et al. 1997a; Davies-Colley and Quinn, 1998), but 90% shading is needed to prevent growth of some emergent macrophytes in low gradient streams (Wilcock et al. 1998).

Shade control of instream primary production also reduces the instream processing of nutrients (uptake of dissolved nutrients into plant biomass) (Quinn et al. 1997b), so that increased shade can result in increased export of dissolved nutrients and higher concentrations downstream (Howard-Williams and Pickmere, 1999). Decomposition of leaf litter from riparian trees also results in uptake of dissolved nutrients from the stream water, but this is not expected to compensate for the reduction in uptake by plants under highly shaded conditions (Quinn et al. 2000a). The overall effect of shade from riparian plantings on downstream nutrient concentrations depends on the balance of the increased riparian uptake versus decreased instream uptake. If nutrient concentrations in a downstream receiving water are judged more important than nuisance plants or high temperature in the reach (e.g., if the stream drains to a nutrient-sensitive lake or river reach of high recreational value), then riparian plantings need to be planned and managed to maintain open lighting conditions (>c. 50%) and to retain nutrient removal functions within the riparian zone (e.g., by managing for low-growing, or spaced, deciduous, riparian vegetation). Because of the site-specific, trade-off nature of shade control to enhance instream uptake of dissolved nutrients, this issue was not included in our riparian function assessments during this study. Modelling studies (Parkyn et al. 2001) indicate that riparian protection/planting that starts in the headwaters will result in lower instream dissolved nutrient concentrations, provided that the groundwater interacts with the riparian area, despite the effect of channel shading lowering instream uptake, because riparian uptake processes will dominate.


2.2.7 Input of large wood and leaf litter

Large wood and leaf litter can play important roles in streams as food resources and habitat (Collier and Halliday, 2000; Quinn et al. 2000b). The role of leaf litter and wood depends on the retentiveness of the stream, which decreases with stream size (Webster et al. 1994, 1999) and flooding frequency. Large wood is most stable in smaller streams, especially where the channel width is less than the typical wood piece length, and in low gradient streams that lack the power during floods to transport wood downstream. Large wood can be a key habitat-forming feature, increasing habitat diversity and cover for invertebrates and fish, and often forms the deepest pools (Quinn et al. 1997a). Wood is particularly important as invertebrate habitat in sandy and silty bedded streams (Collier and Halliday, 2000), and wood input generally increases with the wood density in the riparian area. Restoration of wood in streams, by natural recruitment from restored riparian forest, is a much longer term process (several decades to centuries) than restoration of shade (several years to decades, depending on stream size), because it requires time for tree growth and wood recruitment, via processes including bank undercutting, windthrow and fall of dead trees or branches. Wood recruitment increases with riparian buffer width out to about the maximum height of riparian trees (typically 20-30 m but up to 50 m for large podocarps), beyond which trees are only likely to contribute wood through land slides that enter the channel. However, trees growing closest to the channel contribute the most wood, because they are most likely to drop wood, or fall, into the channel.

2.2.8 Enhancing instream fish habitat and fish spawning areas

Riparian vegetation enhances fish habitat by providing cover and also encourages the input of terrestrial insect food items from overhanging vegetation (Main and Lyon, 1988; Jowett et al. 1996). Cover can take the form of overhanging plants, tree roots, wood and leaf accumulations. Higher over-storey vegetation is less effective fish cover than low-growing grasses and shrubs that grow just above stream level or hang into the stream (pers. comm. R Allibone).

Riparian zones also provide spawning areas for some galaxiid fish species, such as banded kokopu (Mitchell and Penlington, 1982), and short-jawed kokopu that spawn in leaf litter and woody debris during high flows (pers. comm. R Allibone), and inanga that spawn in riparian grasses in tidal lowland reaches (near the salt wedge) (Mitchell and Eldon, 1991). Removal of riparian vegetation in upland areas is expected to reduce the suitability for banded kokopu spawning by eliminating the moist microclimate and leaf litter found under forest, but details of spawning requirements are sketchy. Intensive stock grazing is expected to reduce inanga spawning success by removal of dense grassy vegetation, trampling of eggs and exposure of eggs to desiccation from sunlight and wind during their month-long incubation period.


2.2.9 Controlling downstream flooding

Riparian forest and wetlands are expected to attenuate the peak flow of runoff into the stream channel in small rainfall events (Smith, 1992). Furthermore, well-developed riparian vegetation has greater hydraulic roughness than short grass and hence retards the progress of flood flows as they spill out into the riparian area (Coon 1998, Darby 1999). This water retention may cause increased local flooding of the riparian area and adjacent land, but is expected to reduce the peak flow in downstream reaches. Factors expected to influence the ability of riparian management to control downstream flooding are: the likelihood of overbank flows (less in deeply incised channels); the size of the riparian area and floodplain; the extent of wetlands; and the roughness (stem height in relation to the flow depth, stem diameter, stem spacing, and resistance to flattening) of the riparian vegetation (Darby 1999).

2.2.10 Human recreation

Riparian management can influence human recreation of the riparian area and the stream by changing stream aesthetics, naturalness, access, and the fishability of the stream (Mosley, 1989). These effects are generally more important along medium-sized streams, with access to safe swimming and fishing spots, and in areas of high human access, such as urban streams and reserves.

Riparian management also influences boating and canoeing. Overhanging willows and large wood can be hazardous for boating, whereas native planting plays a particularly important role in enhancing recreational use. Walkways, picnicking facilities (tables and seating), weed control (especially blackberry and other invasives) and vehicle parking areas are all important for enhancing recreational use. Angling use requires particular attention to riparian planting design to provide both overhanging cover and low vegetation to allow fly casting.

2.2.11 Landscape and stream aesthetics

Riparian areas can enhance landscape aesthetics substantially by providing vegetation diversity with ribbons of green within developed pastoral and urban landscapes (Mosley, 1989). We have assumed that shrubs and trees have greater aesthetic appeal than grasses, and that native vegetation has more appeal than exotic vegetation.


3. Results

3.1 General characteristics of the Canterbury study sites

The 313 sites included in the survey covered a wide range of conditions from small, 1st order, headwater streams to large, 7th order, braided rivers (Table 2). The typical (median) stream had no riparian fencing, 80% of the 0.6 m high, stream banks were stable, and stock damage to the banks was judged to be minor. Woody debris was typically sparse and, despite low levels of stream shade (median shade ratio = 0.7), periphyton was only “slippery” and macrophytes were typically absent. One side of the stream was 100% fenced at 21% of sites and both sides at 12% of sites. Riparian wetlands were observed at only 10% of the sites, probably reflecting the low rainfall and permeable soils in the study area. The Land Environment of New Zealand (LENZ = Environmental Domains) database indicates the typically sandy soils of plains are usually slightly impeded to moderately well drained, and the sand-gravel soils of the uplands have moderate to good drainage. Tile drains were not observed at any site but 10% of sites had land drainage channels in their vicinity. Dry channels occurred at 15% of the sites.

Table 2: Summary of stream and riparian characteristics (*see Table 1 for index definitions).

Attribute Mean Median SD Min Max Catchment mean slope (°) 22.4 22.5 16.4 0.1 57.9 Local land slope (°) 11.9 5.3 12.8 0.1 52.7 Segment elevation (m a.s.l.) 190 123 184 8 906 Segment slope (cm/m) 1.9 1.0 3.2 -0.9 28.3 Stream order 3.2 3.0 1.2 1.0 7.0 Local rainfall (mm) 859 819 208 531 1520 Local evapo-transpiration (mm) 691 695 31 608 778 Local mean air temperature (°C) 10.9 11.0 1.0 7.5 12.8 Catchment area (km2) 54 15 174 0.3 2520 Accumulated flow (m3 s-1) 0.81 0.08 3.50 -0.16 53.55 Water width (m) 3.4 2.0 4.8 0.0 40.0 Nonvegetated channel width (m) 10.1 3.5 21.0 0.3 225 Bankfull width (m) 14.4 6.0 26.7 0.5 225 Valley bottom width index 4.1 4.0 1.7 1.0 6.0 Channel slope index 4.1 4.0 1.7 1.0 6.0 Shade ratio 1.5 0.7 2.4 0.0 22.9 Bank height right bank (m) 0.9 0.7 0.9 0.1 10.0 Bank height left bank (m) 0.9 0.5 0.8 0.0 5.0 Mean bank height (m) 0.9 0.7 0.8 0.1 6.0 % macrophyte cover 12.2 0.0 23.6 0.0 100 Periphyton index 1.5 1.0 1.3 0.0 4.0 Wood index 1.0 1.0 0.9 0.0 3.0 %Stable bank 64.3 80.0 35.2 0.0 100


Attribute Mean Median SD Min Max Stock access left bank index 0.6 1.0 0.49 0.0 1.0 Stock access right bank index 0.6 1.0 0.49 0.0 1.0 Left and right banks stock access index 1.2 2.0 0.91 0.0 2.0 Stock bank damage index 0.8 1.0 0.87 0.0 3.0 Local slope length index 2.2 2.0 1.19 1.0 4.0 Local land slope (°) 9.7 3.5 9.4 1.0 30 Local land slope index 3.0 2.0 1.7 1.0 6.0 Tile drain index 0.0 0.0 0.0 0.0 0.0 Drainage channel index 0.1 0.0 0.29 0.0 1.0 Riparian wetland presence index 0.2 0.0 0.42 0.0 1.0 Waste discharge index 0.02 0.0 0.13 0.0 1.0 Fence type index 1.0 0.0 1.38 0.0 3.00 Fence to left bank stream (m) 12.9 5.0 20.6 0.0 100 Fence to right bank stream (m) 10.5 5.0 17.1 0.0 100 Both sides fenced (%) 11.5 10 0.0 0.0 100 One side fenced (%) 21.1 20 0.0 0.0 100

Most sites were classified by the REC as having pastoral catchments with cool dry climates, low elevation source of flow, alluvium geology, middle order (> low order) network position, and low gradient valley land form (Fig. 2). Annual local rainfall (median 819 mm) ranged from 531 to 1520 mm, but evapo-transpiration was high (median 695 mm) so that the accumulated flow of upstream reaches (calculated from rainfall and evapotranspiration) was negative at some sites (indicating ephemeral flow). The channel was dry at 14.5% of the sites during the surveys.


Catchment landcover

0 20 40 60 80

Pasture

Tussock

Scrub

Native forest

Exotic forest

Urban

Bare

Percentage of sites

Catchment climate

0 20 40 60 80

Cool dry

Cool w et

Warm dry

Percentage of sitesSource of flow

0 20 40 60 80

Low Evelation

Hill

Mountain

Glacial

Lake

Percentage of sites

Geology

0 10 20 30 40 50

Alluvium

Hard sed.

Volc. basic

Soft sed.

Misc.

Percentage of sitesNetwork position

0 20 40 60

Mid-order

Low order

High order

Percentage of sites

Valley landform

0 20 40 60 80

Lowgradient

Mediumgradient

Highgradient

Percentage of sites

Figure 2: Summary of survey sites’ REC characteristics.


Figure 3: Channel form at study sites

Grass was the most common dominant riparian vegetation type (48% of sites), followed by willows (26%), low shrubs (9%), and native trees (8%) (Fig. 4). Conifers and deciduous trees (usually a mix of willow and poplar) were each dominant at 3% of sites. Wetland plants (flax, sedges, rushes) were dominant at only 2% of sites, and bare soil at 0.6%.

Figure 4: Dominant riparian vegetation at Canterbury stream sites.

Dominant riparian vegetation type

0.0

10.0

20.0

30.0

40.0

50.0

gras

s

will

ows

low

shr

ub

nativ

e tre

es

coni

fero

us

deci

duou

s

wet

land

high

shr

ubs

bare

soi

l

euca

lypt

us

Perc

enta

ge o

f site

s Channel form

0 20 40 60 80

channelised

meandering

sinuous

straight

Percentage of sites


3.2 Assessment of riparian functions

The field assessments of riparian functions are summarised in Table 3. The various functions differed in their average assessed current and potential activity and also varied widely in activity between the sites. Denitrification was assessed as currently the least active function, whereas bank stabilisation was judged the most active. Applying best practicable riparian management at the sites was judged to be capable of improving most riparian functions substantially (Table 3). The average improvement in function expected followed the order of shading > input of litter and wood, and bank stabilisation > fish habitat, overland flow, and down-stream flood control > recreation > denitrification.

Table 3: Summary of the assessed current (_C) and potential (_P) riparian functions at sites in 3 sub-regions of Canterbury. Functions scored from 0 (not active) to 5 (very highly active).

Function Mean median SD min Max Current functions Bank stability_C 2.2 2 1.3 0 5 Overland flow filtering_C 1.9 2 1.2 0 5 Nutrient uptake _C 2.1 2 1.2 0 5 Denitrification _C 0.7 0 0.9 0 4 Shade for temp_C 1.7 1 1.5 0 5 Shade for plant control_C 1.7 1 1.5 0 5 Wood input_C 1.5 1 1.3 0 5 Litter input _C 1.6 1 1.3 0 5 Fish habitat_C 1.2 1 1.2 0 5 Downstream flooding_C 1.7 2 1.2 0 5 Recreation_C 1.3 1 1.4 0 5 Aesthetics_C 2.0 2 1.4 0 5 Potential functions Bank stability_P 3.7 4 1.0 0 5 Overland flow_P 3.0 3 1.2 0 5 Nutrient uptake_P 3.4 3 1.1 1 5 Denitrification_P 1.3 1 1.3 0 5 Shade for plants_P 3.5 4 1.5 0 5 Shade for temperature_P 3.6 4 1.5 0 5 Wood input_P 3.2 3 1.1 0 5 Litter input_P 3.0 3 1.2 0 5 Fish habitat_P 2.2 2 1.5 0 5 Downstream flooding_P 2.8 3 1.0 0 5 Recreation_P 1.7 2 1.5 0 5 Aesthetics_P 3.3 3 1.0 0 5


3.3 Predicting riparian function activity

3.3.1 Predicting current riparian functions

The factors influencing the activity of the riparian functions were evaluated using multivariate statistics (MOPED programme developed by Ian Jowett of NIWA). First, the sites with similar current riparian function ratings were grouped together in a 3 x 4 Self Organising Map (SOM) using k-medoids (Kauffman and Rousseeuw 1990). The Silhouette Index indicated that these 12 RMC-C cells formed 2 main RMC-C groups between which each of the current riparian function ratings differed significantly (ANOVA, P < 0.1) (see APPENDIX 1 for site classification details). RMC-C group 1 sites had relatively low function ratings compared with group 2 (Fig. 5). Differences in environmental variables between the RMC-C groups 1 and 2 were tested by one-way ANOVA. Six variables assessed on site during the surveys differed most strongly between the groups (Table 4). These were related to dominant riparian vegetation type, shade ratio, stream water width, stock damage to the stream banks, % stable stream bank, and the length and angle of the local land slope and channel slope. Several variables derived from the River Environment Classification (REC) and LENZ databases also showed statistically significant differences between RMC-C groups 1 and 2 (see Table 4).

Table 4: Results of one way ANOVA of environmental variables amongst 2 groups of sites based on current riparian functions. * = variables were log transformed with averages reported as geometric means, other variables are arithmetic means. Results shown for variables with ANOVA P < 0.1 only.

RMC-C ANOVA F P 1 2 Number of sites 175 137 Dominant riparian veg type 199.9 <0.0001 1.93 5.21 Shade ratio* 58 <0.0001 0.41 1.26 stock Bank damage Index 34.2 <0.0001 1.01 0.46 %Stable bank 14.1 <0.0001 57.90 72.66 SLOPE_SECT 13.4 <0.0001 1.38 2.67 Local land slope length index 13.2 <0.0001 2.02 2.50 Loc Land Slope Class 9.2 0.003 2.71 3.31 CTCHSLOPE 8.9 0.003 10.00 14.31 Channel Slope Index 8.7 0.004 2.53 2.96 PSIZE 8.6 0.004 2.09 2.50 ACCSLOPE 7.3 0.007 20.09 25.08 Local Land Slope Metrics 7.1 0.008 8.47 11.29 Log nonvegetated width* 7 0.009 3.32 4.82 INDURATION 5.8 0.016 2.82 3.10 Water Width +1* 5.6 0.018 2.88 3.54 CTCHRAIN 4.7 0.03 836 888 Mean Bank Ht* 4.4 0.037 0.64 0.76 Geol index 4.2 0.041 2.91 3.26


RMC-C ANOVA F P 1 2 SOF Index 3.6 0.058 1.36 1.48 Bankfull width* 3.5 0.062 6.27 7.89 Land cover index 2.9 0.086 3.47 3.70

Figure 5: Comparison of mean (+SE) current function ratings amongst the individual Self Organising Map (SOM) cells (c 1 - c12) and the two major groups based on current riparian function ratings at 313 Canterbury stream sites. 0 = absent, 1 = very low activity, 2 = low-moderate activity, 3 = moderate activity, 4 = high activity and 5 = very high activity.

Group 1 sites (n = 175, SOM cells 1, 2, 5, 6, 9 and 10) had low riparian function ratings (mean ± S.D. of 12 functions = 0.9 ± 0.5, Fig. 5). These were typically unshaded streams, with grass as the dominant riparian vegetation, stock access to one or both banks, and low bank stability (Table 4). Group 2 sites (n = 137, SOM cells 3, 4, 7, 8, 11, and 12) had moderate to high current riparian function ratings (mean ± S.D. of 12 functions = 2.5 ± 0.7,

Litter-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Bank stability-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Overland flow-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Denitrification-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Shade for temperature-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Shade for temperature-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Wood input-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Fish habitat-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12 Gp 1 Gp 2

Recreation-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Aesthetics-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Nutrient uptake-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Downstream flooding-C

0

1

2

3

4

5

c1 c2 c5 c6 c9 c10 c3 c4 c7 c8 c11 c12

Gp 1 Gp 2

Cur

rent

Rip

aria

n Fu

nctio

n R

atin

g


Fig. 5). These sites typically had trees as the dominant riparian vegetation, more stream shade, higher land and channel slopes and slightly wider channels. Variation in functions was greater at the 12 cell classification level (Fig. 5), with the mean rating of all 12 functions ranging from 0.4 ± 0.2 for cell 1 to 3.3 ± 0.5 for cell 12. Compared to cell 12 sites (n = 43; e.g., Photo 2), cell 1 sites (n = 64; e.g., Photo 1) were much wider (mean 13 m wide non-vegetated channel c.f. 3.5 m), poorly shaded (shade ratio = 0.4 c.f. 4.4), and more likely to be dry (31% c.f. 7% of sites).

Photos 1 and 2: Examples of RMC-C cell 1 (group 1) (upper photo) and cell 12 (group 2) sites.


The variables in Table 4 were used to develop a discriminant model to predict SOM cell and group affinities. The SOM group (1 and 2) model (see spreadsheet RMCmodelequations.xls on CD for models) assigned 83% of the sites to the correct group (c.f. 50% expected by chance).

The 83% correct prediction rate for site group classification indicates that this discriminant model could be used as a way for non-experts to assess likely riparian function ratings at a site, based on the mix of site and GIS information in Table 4. A similar model to predict membership amongst the 12 SOM cells based on available GIS and site data had a 57% success rate (c.f. 8% by chance), with an additional 23% of sites being “near misses”, classed in the next most probable cell. These models could be used by people who are not sufficiently trained to make direct onsite evaluations of riparian functions or as a check against individual site evaluation. The models would predict the group or cells affinity and attributes of the site could then be deduced from the typical riparian function ratings for that RMC-C group or cell.

Discriminant models were also developed, using only GIS variables in Table 4 that differed between the RMC-C classes at P < 0.10 information available from REC and LENZ databases. These models assigned 63% of the sites to the correct group (c.f. 50% by chance, see spreadsheet RMCmodelequations.xls for models) and 26% to the correct cell (c.f. 8% by chance). The low hit rate of these models indicates that the currently available GIS data are not suitable for predicting riparian classes and hence riparian functions. The results of these two modelling exercises indicate that reliable prediction of current riparian functions at the regional scale will require additional information on riparian vegetation/shade that is not currently included in the REC or LENZ databases, perhaps from remote sensing using aerial photography of satellite imagery data at levels of resolution appropriate for determining near-stream attributes.

3.3.2 Predicting potential riparian functions

Similar clustering and modelling procedures to those carried out on current riparian functions were also carried out using the potential riparian function activity ratings. This resulted in 3 main RMC-P groups (Fig. 6, Photos 3-5), amongst which each of the riparian functions differed in potential activity ratings (ANOVA, P < 0.05, except for recreation enhancement that was marginally significant P = 0.063) (see APPENDIX 1 for site classification details). Group 2 (n=153, e.g., Photo 4) had the highest average rating for all potential functions (mean ± S.D. for all 12 functions = 3.5 ± 0.4). Typical streams in this group were third order with lowland source of flow and had the highest local land slope (Table 5). Group 3 sites (n = 90, e.g., Photo 5) had the lowest average potential ratings for


shading, litter and wood input, denitrification, nutrient uptake and overland flow filtering (mean ± S.D. for all 12 functions = 2.1 ± 0.5). Typical sites in this group were fourth order, sourced from hill catchment, had large areas, high average catchment slope and high site elevation. Group 1 (n = 69, e.g., Photo 3) had high potential riparian function ratings for shading but low ratings for recreation and fish habitat enhancement. Typical sites in group 1 were third order, with small catchments, lowland source of flow, sand geology, and low catchment and stream channel slope. Thirty-six percent of the sites were dry.

Photos 3, 4 and 5: Examples of RMC-P Group 1 (Cell 1, site E91), Group 2 (Cell 8, site K6), and Group 3 (Cell 9, site K5).


The average potential rating for all 12 riparian functions within the RMC-P cells ranged from 1.8 ± 0.4 for cell 9 (that had the highest average width and 22% with dry channels) to 3.9 ± 0.3 for cell 12 (small streams with the greatest average local land slope). This level of classification is likely to be more useful for finer scale planning.

Figure 6: Comparison of mean potential function ratings (0 = not activity to 5 = very highly active) amongst the individual Self Organising Map (SOM) cells (c1 - c12) and the three major groups based on these ratings at 313 Canterbury stream sites.

Overland flow

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Bank stability

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Denitrification

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Shade for plants

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Gp. 1 Gp. 2 Gp. 3

Nutrient uptake

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Gp. 1 Gp. 2 Gp. 3

Shade for temperature

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Wood input

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Gp. 1 Gp. 2 Gp. 3

Aesthetics

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Recreation

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10Gp. 1 Gp. 2 Gp. 3

Litter input

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Dow nstream flooding

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Gp. 1 Gp. 2 Gp. 3

Fish habitat

0

1

2

3

4

5

c1 c2 c3 c4 c7 c8 c11 c12 c5 c6 c9 c10

Pote

ntia

l Rip

aria

n Fu

nctio

n R

atin

g


Table 5: ANOVA results and mean values of GIS environmental variables in REC and LENZ databases amongst 3 clusters of sites based on potential riparian functions showing average values (*are geometric means).

Anova results SOM Group

Variable F P 1 2 3 Acc Flow* (m3s-1) 60.4 <0.001 0.07 0.12 0.75 SOF Index 38.8 <0.001 1.25 1.26 1.81 Average catchment slope (°) 31.4 <0.001 12.86 21.10 31.51 Catchment area* (km2) 30.7 <0.001 8.55 11.02 41.21 REC land cover index 26.7 <0.001 3.20 3.33 4.27 Stream order 25.8 <0.001 2.70 2.97 3.84 Reach elevation (m) 16 <0.001 188 141 273 Domain particle size class 15.8 <0.001 1.94 2.07 2.87 Local average air temp (°C) 15 <0.001 10.64 11.17 10.57 Domain induration class 12.1 <0.001 2.78 2.77 3.38 DRAINAGE 11.2 <0.001 3.80 3.72 4.38 REC local land slope (°) 9 <0.001 6.32 13.90 12.74 Local evaporation (mm) 8.2 <0.001 678 695 694 Geology index 7.4 0.001 2.57 3.05 3.47 Local annual rainfall (mm) 6.8 0.001 785 865 905 Domain Ca class 3.9 0.022 2.01 1.94 1.83 REC reach slope (cm/m) 3.7 0.026 1.22 2.40 1.72 Local catchment flow (m3 s-1) 2.8 0.066 0.00 0.00 0.01 Domain soil age class 1.3 0.294 1.86 1.78 1.76 Chemical limitation on plants class

1.2 0.326 1.03 1.02 1.00

Domain acid P class 0.7 0.496 3.52 3.56 3.62 Sinuosity 0.7 0.505 1.17 1.19 1.18

A discriminant model (see supplied spreadsheet RMCmodelequations.xls on CD for model) that included all the GIS variables in Table 5 allocated 69% of sites to the correct SOM group (c.f., 33% expected by chance). Function 1 of the discriminant model accounted for 76% of the overall variance explained and was most strongly correlated (canonical structure coefficients) with the log of the calculated flow (Acc Flow; r = 0.67), SOF Index (r = 0.54) and log of catchment area (r = 0.48). Function 2 (24% of variance explained) was most strongly correlated with the local average air temperature (r = 0.46) and the slope of the land draining directly to the stream segment (r = 0.44).

A second discriminant model was developed to predict the SOM cell (n=12, see Fig. 6) affinity using the GIS variables in Table 5 (see supplied spreadsheet RMCmodelequations.xls for model and the GIS layer of predictions of SOM cell RMC-P predictions of each REC reach in Canterbury). This assigned 48% of sites to the correct cell (c.f. about 8% by chance) and 18% of classification were “near misses” (assigned to the


next most probable cell), so that overall 66% of classifications were correct or near misses. Similarly to the model for the 3 SOM groups, function 1 of the SOM cell model (37% of variance explained) was most strongly correlated with log of the calculated flow (r = 0.72) and log of catchment area (r = 0.53), and function 2 (26% of variance explained) was most strongly correlated with the local average air temperature (r = 0.46) and the slope of the land draining directly to the stream segment (r = 0.49).

The better performance of the discriminant function models for predicting potential than current riparian management classes from GIS variables reflects the strong influence on the current function ratings of the current local land management, that is not currently dealt with in the available GIS databases.

These discriminant models provide a means for non-experts to assess the likely potential riparian functions at a site by inputting key information on the site characteristics into the model and examining the characteristics of the cluster to which the site is allocated. This should improve the basis for deciding: (1) whether riparian management is likely to improve functions that provide benefits to the local of downstream aquatic ecosystem; (2) what functions can be enhanced and hence the type of riparian management to put in place and (3) the relative priority of sites in a catchment or region for riparian management. The models can also be used to map the distribution of RMC-P classes using information in the River Environment Classification and LENZ GIS databases (Figure 7). Note however, that the classification error rates of the models (e.g., average 31% for potential function groups based solely on the GIS databases) mean that the predictions will only be indicative. Nevertheless, they are expected to be useful for broad-scale planning purposes.

LENZ data were not available for 4.7% of REC cells in Canterbury. The reaches affected were mainly on large braided rivers that generally have low potential riparian function ratings. Two options were considered for dealing with this missing data issue: (1) running with predictive models that used only the REC data or (2) using the models that incorporate both REC and LENZ data but not classifying the small percentage of reaches that lacked LENZ data. Option 1 was evaluated by comparing the percentage of correct RMC-P cell and group classifications with and without inclusion of the LENZ variables. Excluding these reduced the % correct group classification from 69% to 64% and correct cell classification from 48% to 39%. From this it was decided to use the models that included REC and LENZ data, because (1) the RMC-P cell prediction accuracy dropped significantly without the LENZ variables (2) a small proportion of reaches would be left unclassified (due to lack of LENZ data), and (3) the unclassified reaches were typically on large braided rivers where we would expect relatively low riparian function potential.


Figure 7: Map showing predicted potential riparian management classification (RMC-P) groups (1-3, upper map) and cells (1-12, lower map) for the coastal area north of Christchurch.

3.3.3 Predicting individual riparian functions

Discriminant function modelling was also used to predict the potential ratings of individual riparian functions (0 to 5) based on GIS variables (see supplied spreadsheet RMCmodelequations.xls for model and GIS layers showing the predictions for each REC


reach in Canterbury). The predictive accuracy of the models and most important GIS environmental predictors (i.e., those with strongest correlations with first 2 functions) are summarised in Table 6. The models allocated 37 – 53% of sites to the correct class and 69-91% of the predictions were either correct or near misses (predicted an adjacent rating, i.e., 3 or 5 if the actual assessed rating was 4). This suggests that these models would be useful for predicting the potential roles of various individual riparian functions at particular locations or mapping variations in potential functions. An example of the mapped prediction for the potential shade for temperature function rating is shown in Figure 8.

Figure 8: Map showing predicted potential riparian shading for instream temperature control function ratings for the coastal area north of Christchurch.


Table 6: Summary of multiple discriminant Analysis modelling to predict potential ratings of individual riparian functions from GIS variables.

Riparian functions % pred.

correct

% near misses

Top correlations: function 1 Top correlations: function 2

Bank stability 41 33 Land cover index; SOF index Local rain, flow

Overland flow 46 33 Local land slope; channel slope Chemical limitation on plants; local temperature

Nutrient uptake 45 43 Local land slope; channel slope SOF index

Denitrification 40 27 Flow, SOF index, induration Soil particle size

Shade for plants 53 28 Flow, upstream catchment area Local rain, local land slope, local evaporation

Shade for temperature

53 27 Flow, upstream catchment area Local rain, local land slope, local evaporation

Wood input 42 33 Local temp, site elevation Chemical limitation on plants

Litter input 41 28 Local temp, channel slope SOF index, flow

Fish habitat 42 28 Local temp, land drainage Local land slope, local evaporation

Downstream flood control

40 28 Local temp, soil phosphorus Land slope of whole catchment

Recreation 47 25 Catchment slope, order Channel slope

Aesthetics 37 32 Local land slope, channel slope SOF index, catchment slope

3.3.4 Geomorphic approach to riparian management classification

The results of the assessments of current and potential riparian functions amongst Canterbury streams highlight some key morphological factors influencing function ratings that need to be considered in management decisions. These were channel width, permanence of flow and the slope of land adjacent to the stream/riparian area.

3.3.4.1 Permanence of flow

The permanence of flow at a site has obvious influences on its values for recreation, and aesthetics. Although knowledge of the ecological roles and values of intermittent streams is rudimentary, I have assumed for the purposes of this exercise that riparian functions to protect local instream values are less important at sites that are usually dry, or dry up during summer. Our on-site assessments of potential riparian functions indicate that the 46 sites (15% of total) that were dry during our spring-summer surveys are similar (i.e., mean rating within 1 unit) to the perennial sites in their potential bank stabilisation, overland


flow control, nutrient uptake from groundwater and downstream flood control functions, but were less active for all other functions. Nevertheless, the seasonal importance of ephemeral reaches for fish spawning (e.g., trout spawning during winter) may result in some ephemeral reaches warranting high priority for riparian management to enhance instream habitat even though the reach is dry during summer. Further research is required on the natural values and roles of ephemeral streams for maintaining the health of downstream ecosystems in order to provide a better basis for their management.

Riparian management planning should, therefore, consider the permanence of flow at sites. Unfortunately, attempts to predict whether sites were dry or not during our spring-summer surveys by discriminant function modelling using the GIS variables were not particularly successful. Stream order, catchment area and predicted flow were all significantly lower at the dry sites, but the discriminant model assigned only 71% of sites to the correct wet/dry class (c.f. 50% by chance). Further work to develop better predictions of flow permanence would improve the basis for riparian management decisions.

3.3.4.2 Channel width

Channel width has a strong influence on the interaction of riparian vegetation and instream habitat (i.e., shading, delivery and retention of wood and leaf litter from the riparian vegetation to the wetted channel, and influence on fish habitat). This suggests that classifying streams by channel width will improve the effectiveness of riparian management.

Our assessments of potential shading function for control of stream temperature and instream vegetation in relation to channel width indicate that the shading function decreases from “high activity” (rating 4-5) to “low-moderate activity” (rating 2) at a non-vegetated channel width of approximately 10-12 m (e.g., Fig. 9). (Note that some narrower channels have low potential shade ratings because only low growing plants are likely to grow as riparian vegetation (e.g., in tussock areas of the intermontane basins) or because the channels are typically dry, which reduced the shade function rating). This is consistent with changes in stream lighting measured with canopy analysers in relation to stream width and riparian vegetation (Davies-Colley and Quinn, 1998), and indicates that a non-vegetated channel width of 12 m is an appropriate cut-off for distinguishing sites above which the shading functions (temperature and algae control) of riparian trees are likely to be ineffective.

Some inland areas of Canterbury are expected to be unsuitable for growth of riparian trees due to climatic and soil constraints, and natural riparian vegetation is limited to tussock


grasses and small shrubs (e.g., matagouri). This suggests that further subdivision of sites by channel width would be useful for guiding riparian planning. With this in mind, 4 channel width classes are suggested related to the types of riparian vegetation required to provide stream shade: <2 m (T = “tiny”), 2 to <6 m (S = “small”); 6 to <12 m (M = “medium”); and ≥12 m (L = “large). Long pasture grasses and tussocks are expected to shade tiny streams effectively, whereas high shrubs will shade tiny and small streams and trees will shade channels up to the medium size class.

Potential riparian function ratings varied between sites when grouped by these size classes (tested by ANOVA with post-hoc Scheffe multiple comparisons between groups). As expected, potential shade ratings for instream plant control and temperature were highest in the tiny and small channels > medium > large. The large class also has significantly lower potential function ratings for denitrification, nutrient uptake from groundwater, litter and wood input, fish habitat and downstream flooding. Tiny streams had significantly lower average ratings for aesthetics and recreation.

Figure 9: Effect of stream channel width on assessed potential riparian shading function for temperature control at sites in Canterbury. Function ratings range from 0 (no activity) to 5 (very high activity).

0

1

2

3

4

5

0 50 100 150 200 250

Pote

ntia

l sha

de fo

r tem

pera

ture

ratin

g

Non-vegetated channel width (m)


3.3.4.3 Local landform

Local landform is another key morphological influence on riparian functions and management, through its influence on surface runoff. This will generally increase (and buffer widths will need to widen to provide optimal efficiency) as the slope length, angle and clay content of the adjacent land increase and as soil drainage decreases. Collier et al. (1995) provide a method for determining how combinations of these factors influence the optimal width of a buffer strip for controlling sediment in surface runoff. The combination of slope length and angle is expected to increase the optimal width of riparian areas for contaminant filtering as the adjacent landform changes from plain to U-shaped to V-shaped. Our on-site assessments of potential riparian functions amongst sites classified as plain (P), U-shaped or V-shaped showed that riparian areas in V-shaped valleys had ≥1 unit higher average ratings than those in plain areas (Scheffe tests, P < 0.05) for control of overland flow (means 2.6, 3.5 and 4.2 for P, U- and V-shaped landforms, respectively) and recreation functions (means 1.5, 2.3, and 2.5, respectively). Landform also influenced potential ratings for nutrient uptake from shallow groundwater by riparian vegetation (means 3.2, 3.9 and 4 for plain, U and V morphologies), but potential ratings for the other riparian functions did not differ markedly by landform class.

3.3.4.1 Riparian/stream geomorphic classes

Based on the above, the sites were grouped into 12 geomorphic classes using combinations of channel width and valley shape (i.e., PT = Plain/Tiny; PS = Plain/Small; PM = Plain/Medium; PL = Plain/Large; UT = U-shaped/Tiny; US = U-shaped /Small; UM = U-shaped /Medium; UL = U-shaped /Large; VT = V-shaped/Tiny; VS = V-shaped /Small; VM = V-shaped /Medium; VL = V-shaped /Large) (see Photos 6-17 for site examples and Appendix 2 examples of cross-section sketches). Figure 10 shows that all 12 riparian functions showed statistically significant differences in their assessed potential activity amongst these classes. Riparian/stream geomorphology influences what riparian management can deliver to improve stream values and meet defined catchment or site goals. Differences in function ratings were strong (high F statistics) for shading functions (decreasing with size across shape types) > overland flow filtering (V>U>P) > recreation (very low at tiny plain and v-shaped sites) and weakest for downstream flood control. Bank stabilisation and aesthetic enhancement were rated as moderately to very highly active potential functions (mean ≥3) in all classes. Riparian denitrification potential was lowly rated in all classes, but particularly so around large channels. A discriminant function model, using 20 GIS variables that differed significantly between the groups (at P <0.1; all except CHEMLIM and sinuosity), allocated 60% of sites to the correct class (c.f. 8% by chance) and another 17% were classified to an adjacent class. This indicates that the GIS-


based model is useful for preliminary mapping riparian sites into these landform/size classes. An example of the mapped geomorphic classes for an area of Canterbury is shown in Figure 11 (see supplied spreadsheet RMCmodelequations.xls for model and GIS layers showing the predicted geomorphic classes for each REC reach in Canterbury).

Photos 6-9: Examples of Plains Tiny (PT), Small (PS), Medium (PM), and Large (PL) riparian/stream geomorphic classes.


Photos 10-13: Examples of U-shaped valley Tiny (UT), Small (US), Medium (UM), and Large (UL) riparian/stream geomorphic classes.

Photos 14-17: Examples of V-shaped Tiny (VT), Small (VS), Medium (VM) and Large (VL) geomorphic classes.


Figure 10: Average potential ratings (+ SE) for 12 riparian functions amongst twelve geomorphology based classes (RMC-G) at 313 Canterbury stream sites. Landform/width codes: P = plain/floodplain; U = U-shaped; V = V-shaped. Size codes: T = tiny (channel width < 2m); S = small (2 - <6m); M = Medium (6 - <12m); L = Large (≥12 m).

Overland f low f iltering

0

1

2

3

4

5

PT PS PM PL UT US UM UL VT VS VM VL

F = 11.5, p < 0.0001

Bank stabilisation

0

1

2

3

4

5


F = 2.8; p = 0.0017Nutrient uptake from groundw ater

0

1

2

3

4

5


F = 5.8; p < 0.0001

Denitrif ication

0

1

2

3

4

5


F = 4.0; p < 0.0001

Shade for instream plant control

0

1

2

3

4

5


F = 17.3; p < 0.0001

Shade for stream temperature control

0

1

2

3

4

5


F = 20.4; p < 0.0001

Wood input

0

1

2

3

4

5


F = 2.5; p < 0.0048

Leaf litter input

0

1

2

3

4

5


F = 4.9; p < 0.0001

Fish habitat

0

1

2

3

4

5


F = 2.6; p < 0.0031

Dow nstream f lood control

0

1

2

3

4

5


F = 2.1; p = 0.02

Recreation

0

1

2

3

4

5


F = 9.8; p < 0.0001

Aesthetics

0

1

2

3

4

5


F = 2.9; p < 0.0031

Pot

entia

l Rip

aria

n Fu

nctio

n R

atin

g


Figure 11: Map showing predicted geomorphology based classes (RMC-G) for the coastal area north of Christchurch. Landform/width codes: P = plain/floodplain; U = U-shaped; V = V-shaped. Size codes: T = tiny (channel width < 2m); S = small (2 - <6m); M = Medium (6 - <12m); L = Large (≥12 m).

3.3.4.2 Other potential classification attributes

High rates of riparian denitrification require groundwater flows through water-logged soils (anoxic conditions) with a source organic carbon (Barton et al. 1999). Recent research in the USA has used soil map data to identify area with wet soils as a planning tool for riparian management to enhance denitrification (Gold et al. 2001). In Canterbury, heavy, slow draining soils (where denitrification is expected to be operative in riparian zones) have been mapped (Main, 2003 (in prep.)) from data in Kear et al. (1967), and this map (Fig. 12) will be useful for identifying areas where riparian zones can probably be managed effectively for denitrifying inflows of shallow groundwater to streams. Free-draining soils, that cover approximately 77% of the Canterbury Plains (mostly Lismore series), are also unlikely to generate much surface runoff whereas the opposite is expected for heavy soils. Thus, overlaying the map of heavy, poorly drained soils should also identify the areas where riparian management is likely to be needed to deal with surface runoff in flatter areas and also where denitrification has potential to reduce nitrate inputs from shallow groundwater.


N

EW

S

70000 0 70000 140000 Meters

Figure 12: Extent of heavy, slow draining soils on the Canterbury plains as listed by Kear et al. (1967).


3.4 Application of riparian management classification in river management

The preceding sections outline the development of approaches to riparian management classification for Canterbury streams. Figure 13 provides a proposed framework for using RMC in catchment management.

Once specific goals are established for a river at the catchment and/or segment scale (step 1), predictions of current (RMC-C) and potential RMC classes (RMC-P and/or geomorphic classes), or potential ratings of specific riparian attributes, can be made using the discriminant function models. These classes indicate the riparian functions that currently contribute to the river management goals (from RMC-C class and mean function activity ratings in Fig. 5) and the potential functions that can be enhanced by riparian management (predicted RMC-P class and mean function ratings in Figs. 6 and/or geomorphic class and mean potential functions in Fig. 10) (steps 2 and 3 in Fig. 13).

Because of the need for field data for reasonably reliable predictions of RMC-C classes, this step will involve surveys of a representative sample of site and extrapolation to similar sites. The information in Sections 2.2 and 2.3 provides a basis for direct on-site evaluation of current and potential riparian functions that can be checked against the predictions of discriminant function models. The riparian function information from the RMC-C, RMC-P and geomorphic classifications provides an improved basis for prioritising areas for riparian management to meet the river goals (step 4). This feeds into riparian management strategies, which also recognise other goals, pressures, and the available resources (Step 5), and provide the context for reach or farm scale riparian management plans by/with farmers and other land-owners (step 6). Finally, the riparian microhabitat-based native plant recommendations (see Fig. 7 and Table 9 in Quinn et al. 2001) feed into this step by providing the detail needed to improve the success of native plantings.


Figure 13: Flow chart showing how Riparian Management Classification (RMC) and microhabitat based native plant recommendations can contribute to river management planning.

5. Acknowledgements

Thanks to staff of Environment Canterbury (particularly Cathie Brumley) for support and discussions during this study. The field surveys were carried out by Rochelle Lavender (ECan) and Alastair Suren (NIWA). Kerrie Niven, Helen Hurren, and Mark Weatherhead, NIWA, provided excellent GIS support. John Leathwick and his team provided the data from the Land Environments of New Zealand (LENZ) database. Ruth Berry and Ton Snelder helped with the formulation of this project. The report benefited from constructive reviews by Stephanie Parkyn and Kevin Collier.

Riparian Management Classification

River management1. Identify specific goals for catchment and river segments

2. Identify how riparian functions contribute to goals currently

Evaluate current RMC classes from field survey and map information using discriminant models and infer riparian function ratings (Fig. 2) or make direct assessments

3. Evaluate potential riparian function ratings with best practicable riparian management

Evaluate potential RMC classes from field survey and map information using discriminant models and infer function ratings (Fig. 3) or make direct assessments

4. Identify priority river segments for riparian management where riparian functions that contribute to goals are predicted to increase most in activity

5. Develop riparian management strategies recognising variations in priorities and riparian functions within the catchment or region

Other relevant informatione.g. Statutory obligations, Maori perspectives, local interest/politics, terrestrial biodiversity goals, landscape ecology issues and available resources

6. Reach and farm scale riparian management plans

Farmer/landowners goals for their properties

Riparian microhabitat-based native species planting recommendations


6. References

Abernathy, B.; Rutherfurd, I.D. (1999). Guidelines for stabilising streambanks with riparian vegetation. Parkville, Victoria, Cooperative Research Centre for Catchment Hydrology.

Barton, L.; McLay, C.D.A.; Schipper, L.A.; Smith, C.T. (1999). Annual denitrification rates in agricultural and forest soils: a review. Australian Journal of Soil Research 37, 1073-1093.

Biggs, B.J.F. (2000). New Zealand periphyton guideline: detecting, monitoring and managing enrichment of streams. Ministry for the Environment.

Boothroyd, I.K.G.; Langer, E.R. (1999). Forest harvesting and riparian management: A review. Wellington, NIWA.

Brown, P.; Mackay, S. (2000). Case study - Riparian management on the Piako River: A new approach to costs and benefits. Hamilton, Environment Waikato.

Burckhardt, J.C.; Todd, B.L. (1998). Riparian forest effect on lateral stream channel migration in the glacial till plains. Journal of the American Water Resources Association 34, 179-184.

Collier, K.J.; Cooper, A.B.; Davies-Colley, R.J.; Rutherford, J.C.; Smith, C.M.; Williamson, R.B. (1995). Managing riparian zones: A contribution to protecting New Zealand's rivers and streams. vol. 1: Concepts. Department of Conservation, Wellington, NZ.

Collier, K.J.; Halliday, J.N. (2000). Macroinvertebrate-wood associations during decay of plantation pine in New Zealand pumice-bed streams: stable habitat or trophic subsidy? Journal of the North American Benthological Society 19, 94-111.

Coon, W.F. (1998). Estimation of roughness coefficients for natural stream channels with vegetated banks. US. Geological Survey water-supply paper 2441.

Cooper, A.B. 1990. Nitrate depletion in the riparian zone and stream channel of a small headwater catchment. Hydrobiologia 202, 13-26.


Cooper, A.B.; Smith, C.M.; Smith, M.J. (1995). Effects of riparian set-aside on soil characteristics in an agricultural landscape: Implications for nutrient transport and retention. Agriculture, Ecosystems and Environment 55, 61-67.

Darby, S.E. (1999). Effect of riparian vegetation on flow resistance and flood potential. Journal of Hydraulic Engineering-ASCE 125: 443-454.

Davies-Colley, R.J.; Quinn, J.M. (1998). Stream lighting in five regions of North Island, New Zealand: control by channel size and riparian vegetation. New Zealand Journal of Marine and Freshwater Research 32, 591-605.

Dunaway, D.; Swanson, S.R.; Wendel, J.; Clary, W. (1994). The effects of herbaceous plant communities and soil textures on particle erosion of alluvial streambanks. Geomorphology 9, 47-56.

Gold, A.J.; Groffman, P.M.; Addy, K.; Kellogg, D.Q.; Stolt, M.; Rosenblatt, A.E. (2001). Landscape attributes as controls on ground water nitrate removal capacity of riparian zones. Journal of the American Water Resources Association 37, 1457-1464.

Gregory, S.V.; Swanson, F.J.; McKee, W.A.; Cummins, K.W. (1991). An ecosystem perspective on riparian zones. Bioscience 41, 540-551.

Hill, A.R. (1996). Nitrate removal in stream riparian zones. Journal of Environmental Quality 25, 743-755.

Howard-Williams, C.; Pickmere, S. (1999). Nutrient and vegetation changes in a retired pasture stream : recent monitoring in the context of a long-term dataset. Department of Conservation, Wellington.

Jowett, I.G.; Richardson, J.; McDowall, R.M. (1996). Relative effects of in-stream habitat and land use on fish distribution and abundance in tributaries of the Grey River, New Zealand. New Zealand Journal of Marine and Freshwater Research 30, 463-475.

Kauffman, L.; Rousseeuw, P.J. (1990). Finding groups in data: an introduction to cluster analysis. John Wiley and Sons, New York.

Kear, B.S.; Gibbs, H.S.; Miller, R.B. (1967). Soils of the downs and plains of Canterbury and North Otago, DSIR Soil Bureau Bulletin 14, 92 pp.


Knowles, R. (1982). Denitrification. Microbial Reviews 46, 43-70.

Lowrance, R.; Altier, L.S.; Newbold, J.D.; Schnabel, R.R.; Groffman, P.M.; Denver, J.M.; Correll, D.L.; Gilliam, J.W.; Robinson, J.L.; Brinsfield, R.B.; Staver, K.W.; Lucas, W.; Todd, A.H. (1997). Water quality functions of riparian forest buffers in Chesapeake Bay watersheds. Environmental Management 21, 687-712.

Lyons, J.; Trimble, S.W.; Paine, L.K. (2000). Grass versus trees: managing riparian areas to benefit streams of central North America. Journal of the American Water Resources Association 36, 919-930.

Main, M.R.; Lyon, G.L. (1988). Contribution of terrestrial prey to the diet of banded kokopu (Galaxias fasciatus Gray) (Pisces: Galaxiidae) in South Westland, New Zealand. Verhandlungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie 23, 1785-1789.

Main, M.R. (2003) (in prep.). Review of draft setback distance rules in Discussion Draft NRRP Water Quality Chapter 7, with respect to land applications of contaminants and surface waters, Environment Canterbury.

MFE. (2001). Managing waterways on farms: A guide to sustainable water and riparian management in rural New Zealand. Ministry for the Environment, Wellington. 209 p.

Mitchell, C.P.; Penlington, B.P. (1982). Spawning of Galaxias fasciatus Gray (Salmoniformes: Galaxiidae). New Zealand Journal of Marine and Freshwater Research 16, 131-133.

Mitchell, C.P.; Eldon, G.A. (1991). How to locate and protect whitebait spawning grounds. Freshwater Fisheries Centre Publication for the Department of Conservation.

Mosley, M.P. (1989). Perceptions of New Zealand river scenery. New Zealand Geographer 45, 2-13.

Murgatroyd, A.L.; Ternan, J.L. (1983). The impact of afforestation on stream bank erosion and channel form. Earth Surface Processes and Landforms 8, 357-369.

Nguyen, M.L.; Sheath, G.W.; Smith, C.M.; Cooper, A.B. (1998). Impact of cattle treading on hill land. 2. Soil physical properties and contaminant runoff. N.Z. J. Agr. Res. 41, 279-290.


Parkyn, S.; Davies-Colley, R.; Cooper, B.; Collier, K.; Reeves, P.; Stroud, M. (2001). "Downstream downsides to riparian management?" In: Presented at the International Ecological Engineering Conference, Lincoln University, Lincoln, NZ.

Phillips, J.D. (1989a). An evaluation of the factors determining the effectiveness of water quality buffer zones. Journal of Hydrology 107, 133-145.

Phillips, J.D. (1989b). Nonpoint source pollution control effectiveness of riparian forests along a coastal plain river. Journal of Hydrology 110, 221-237.

Prosser, I.; Bunn, S.; Mosisch, T.; Ogden, R.; Karssies, L. (1999). The delivery of sediment and nutrients to streams. In: Price, P.; Lovett, S. Riparian Land Management Technical Guidelines. Land and Water Resources Research and Development Corporation (LWRRDC), Canberra. 1: Principles of Sound Management, pp. 37-60.

Quinn, J.M.; Williamson, R.B.; Smith, R.K.; Vickers, M.L. (1992). Effects of riparian grazing and channelization on streams in Southland, New Zealand. 2. Benthic invertebrates. New Zealand Journal of Marine and Freshwater Research 26, 259-269.

Quinn, J.M.; Cooper, A.B.; Davies-Colley, R.J.; Rutherford, J.C.; Williamson, R.B. (1997a). Land use effects on habitat, water quality, periphyton, and benthic invertebrates in Waikato, New Zealand, hill-country streams. New Zealand Journal of Marine and Freshwater Research 31, 579-597.

Quinn, J.M.; Cooper, A.B.; Stroud, M.J.; Burrell, G.P. (1997b). Shade effects on stream periphyton and invertebrates: an experiment in streamside channels. New Zealand Journal of Marine and Freshwater Research 31, 665-683.

Quinn, J.M. (1999). Towards a riparian zone classification for the Piako and Waihou River catchments.

Quinn, J.M.; Burrell, G.P.; Parkyn, S.M. (2000a). Influence of leaf toughness and nitrogen content on in-stream processing and nutrient uptake in a Waikato, New Zealand, pasture stream and streamside channels. New Zealand Journal of Marine and Freshwater Research 34, 255-274.

Quinn, J.M.; Smith, B.J.; Burrell, G.P.; Parkyn, S.M. (2000b). Leaf litter characteristics affect colonisation by stream invertebrates and growth of Olinga feredayi (Trichoptera: Conoesucidae). New Zealand Journal of Marine and Freshwater Research 34, 275-289.


Quinn, J.M.; Suren, A.M. (2001). Riparian management classification for Canterbury streams. Proceedings of International Ecological Engineering Conference, Lincoln University, Christchurch, NZ.

Quinn, J.M.; Suren, A.M.; Meurk, C.D. (2001). Riparian management classification for Canterbury streams. Hamilton, NIWA Client Report MFE01229/1, 50 p.

Quinn, J.M.; Brown, P.M.; Boyce, W.; Mackay, S.; Taylor, A.; Fenton, T. (2001). Riparian zone classification for management of stream water quality and ecosystem health. Journal of the American Water Resources Association 37(6): 1509-1515.

Rutherford, J.C.; Blackett, S.; Blackett, C.; Saito, L.; Davies-Colley, R.J. (1997). Predicting the effects of shade on water temperature in small streams. New Zealand Journal of Marine and Freshwater Research 31, 707-721.

Rutherford, J.C.; Davies-Colley, R.J.; Quinn, J.M.; Stroud, M.J.; Cooper, A.B. (1999). Stream shade: towards a restoration strategy. Department of Conservation, Wellington, NZ.

Rutherfurd, I.; Abernathy, B.; Prosser, I. (1999). Stream erosion. In: Lovett, S.; Price, P. Riparian land management technique guidelines. Volume one: Principles of sound management. LWRRDC, Canberra, pp. 60-78.

Smith, C.M. (1989). Riparian pasture retirement effects on sediment, phosphorus, and nitrogen in channelised surface run-off from pastures. New Zealand Journal of Marine and Freshwater Research 23, 139-146.

Smith, C.M. (1992). Riparian afforestation effects on water yields and water quality in pasture catchments. Journal of Environmental Quality 21, 237-245.

Snelder, T.; Weatherhead, M.; O'Brien, R.; Shankar, U.; Biggs, B.J.; Mosley, P. (1999). Further development and application of a GIS based river environmental classification system. Christchurch, NIWA Client Report CHC99/01.

Thorne, C.R. (1990). Effects of vegetation on riverbank erosion and stability. In: Thornes, J. B. Vegetation and erosion. Wiley, Chichester.


Watson, A.; Phillips, C.; Marden, M. (1999). Root strength, growth and rates of decay: root reinforcement changes of two tree species and their contribution to hillslope stability. Plant and Soil 217, 39-47.

Webster, J.R.; Benfield, E.F.; Ehrman, T.P.; Schaeffer, M.A.; Tank, J.L.; Hutchens, J.J.; DAngelo, D.J. (1999). What happens to allochthonous material that falls into streams? A synthesis of new and published information from Coweeta. Freshwater Biology 41(4): 687-705.

Webster, J.R.; Covich, A.P.; Tank, J.L.; Crockett, T.V. (1994). Retention of coarse organic particles in streams in the Southern Appalachian mountains. Journal of the North American Benthological Society 13(2): 140-150.

Wilcock, R.J.; Rodda, H.J.E.; Scarsbrook, M.R.; Cooper, A.B.; Stroud, M.J.; Nagels, J.W.; Thorrold, B.S.; O'Connor, M.B.; Singleton, P.L. (1998). The influence of dairying on the freshwater environment (the Toenepi study). Volume 2. Technical report. Hamilton, NIWA.

Willems, H.P.L.; Rotelli, M.D.; Berry, D.F.; Smith, E.P.; Reneau, R.B.; Mostaghimi, S. (1997). Nitrate removal in riparian wetland soils: Effects of flow rate, temperature, nitrate concentration and soil depth. Water Research 31, 841-849.

Williamson, R.B.; Smith, C.M.; Cooper, A.B. (1996). Watershed riparian management and its benefits to a eutrophic lake. Journal of Water Resources Planning and Management 122, 24-32.


7. Appendix 1: List of study site names, locations and riparian management classifications. Classifications are presented according to their current (RMC-C) and potential riparian functions determined statistically (RMC-P) and inferred from knowledge (Geomorphic class: PT = Plain/Tiny; PS = Plain/Small; PM = Plain/Medium; PL = Plain/Large; UT = U-shaped/Tiny; US = U-shaped /Small; UM = U-shaped /Medium; UL = U-shaped /Large; VT = V-shaped/Tiny; VS = V-shaped /Small; VM = V-shaped /Medium; VL = V-shaped /Large).

Site No Location Map

No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

1 Halswell River @ Motukarara M36 772198 9 1 8 2 PM

2 Drain on Whithells Rd M36 785208 2 1 1 1 PS

3 McQueens Ck @ SH75 M36 787196 9 1 2 1 PS

4 Halswell River @ Seabridge Rd M36 776172 1 1 3 2 PM

5 Halswell @ Seabridge Rd, Upper M36 782173 9 1 6 3 PM

6 small stream near Ataahua M36 803164 5 1 2 1 PT

7 Kaituna Rv @ SH75 (Ataahua) M36 852147 9 1 7 2 PM

8 Okana Stream M36 847174 1 1 3 2 PS

9 Kaituna Rv @ Parkinsons Rd M36 854194 1 1 12 2 PM

10 small trib into Kaituna Rv on Parkinsons Rd M36 853195 8 2 12 2 PS

11 trib into Kaituna Rv (upper near road end) M36 900210 12 2 8 2 US

12 Upper Kaituna Valley M36 895212 12 2 12 2 US

13 Mid Prices Stream M36 864153 1 1 8 2 PS

14 Upper Prices Stream M36 876161 3 2 8 2 US

15 Prices Stream near Willesden M36 859143 1 1 8 2 US

16 Prices Stream @ SH75 M36 845126 4 2 4 2 PS

17 Garry Rv on Birch Hill Road M34 570756 7 2 10 3 PL

18 Wooded Gully M34 561793 12 2 12 2 UM

19 Maori Stream M34 533786 4 2 11 2 PL

20 Washpool Stream on Birch Hill Rd M34 535754 10 1 4 2 PS

21 Glentui Rv on Ashley Gorge Rd M34 524756 12 2 3 2 PM

22 Glentui Rv @ DOC picnic site L34 492783 12 2 12 2 VS

23 Ashley Rv @ Road bridge L34 475752 4 2 9 3 VL

24 Ashley Rv @ lower bridge L34 426757 2 1 9 3 VL

25 Trib into Ashley off Ladbrooks Hill L34 417774 12 2 12 2 VS

26 Townsend Stream near Ashley Rv L34 400793 1 1 9 3 UL

27 Ashley Rv above gorge L34 410798 1 1 9 3 UL

28 Five Gully Stream, Mt Pember Station L34 405818 1 1 1 1 PS

29 Whistler Stream L34 412827 1 1 9 3 PL



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

30 Broom Stream L34 435845 1 1 1 1 PS

31 Ashley, Upper road bridge L34 466867 1 1 9 3 PL

32 Duck Creek, on road to Island Hill L34 466876 11 2 7 2 PS

33 Stream draining Ashley (Hill?) L34 478888 1 1 5 3 UM

34 Stream past Okuku Falls M34 525934 10 1 4 2 VT

35 trib into Okuku Downs Stream M34 548938 3 2 3 2 VS

36 Okuku Rv @ upper Ford M34 570948 1 1 5 3 PL

40 Trib into Cam on Church Bush Rd M35 821623 5 1 4 2 PS

41 Cam Rv @ Revells Rd M35 817617 7 2 12 2 PM

42 Cam @ Youngs Rd M35 802634 11 2 8 2 PM

43 Stream on North Brook Rd M35 792663 12 2 4 2 PT

44 Cam Rv @ Camside Rd M35 796660 12 2 4 2 PT

45 Cam @ Coldstream M35 788874 12 2 4 2 PT

46 Southbrook @ Lineside Rd M35 776646 9 1 12 2 PS

47 Stream beside Rangiora/PLaxton Rd M35 776635 12 2 4 2 PT

48 Stream on Todds Rd/Fernside Rd M35 768638 10 1 4 2 PT

49 Stream on North Todd Rd M35 772643 9 1 4 2 PT

50 Southbrook on Townsend Rd M35 763651 12 2 4 2 PS

51 Small Drain Into Townsend Rd M35 764644 1 1 4 2 PT

52 Small stream on No.5 Drain Rd M35 760643 1 1 4 2 PT

53 Small stream on No.5 Drain Rd at Elarish M35 760638 10 1 4 2 PS

54 Stream on Fawcetts Rd (upstream of road) M34 765703 10 1 1 1 PT

55 Stream on Fawcetts Rd (downstream) M34 765703 5 1 1 1 PM

56 Stream on Dixons Rd M34 749711 11 2 1 1 PS

57 Stream on Mowatts Rd M34 743719 11 2 1 1 PS

58 Stream on Carrs Rd near Mowatts M34 745726 11 2 1 1 PS

59 Trib into Makerikeri Rv, Barkers Rd M34 728711 10 1 1 1 PT

60 Stream on Barkers Rd corner M34 687712 11 2 4 2 PT

61 Stream on Barkers Rd near Swamp rd M34 697712 1 1 1 1 PT

62 Waipara Rv @ Stringers Rd M34 830938 4 2 10 3 PL

63 Waipara @ Ladmores Rd M34 766941 4 2 10 3 PL

64 Small trib on Ram Paddock Rd M34 730935 11 2 1 1 VS

65 Forestors Culvert on Ram Paddock Rd M34 728938 12 2 4 2 US

66 Waipara Rv south branch M34 727958 3 2 9 3 PL

67 Dry Creek in Forestry Block M34 725960 1 1 1 1 UT

68 Waipara Rv, North Branch M34 732990 3 2 10 3 UL

69 Reservation gully, Mid Waipara branch M34 685988 11 2 7 2 VS

70 Mid Waipara @ McDonald Downs Rd M34 729990 7 2 11 2 UM



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

71 Tommy's Stream M33 712019 12 2 4 2 PS

72 North Branch Waipara Rv M33 734042 12 2 11 2 PM

73 Waipara Rv Nth Branch @ Heathstock rd M33 726084 1 1 1 1 PM

74 Dyfryn? Stream on Virginia Rd M33 695108 10 1 2 1 VT

75 North Branch Waipara Rv @ Virginia Rd M33 668093 12 2 12 2 VM

76 North Branch Waipara Rv @ Pannetts Bridge M33 635096 9 1 4 2 US

77 Pig Gully M33 615094 5 1 4 2 VT

78 Washpen Stream @ Murrays Rd M33 742128 5 1 4 2 PT

79 Waitohi Rv @ Powers Rd M33 740173 8 2 10 3 PL

80 Washpen Stream @ Horsley Down Rd M33 800120 12 2 8 2 PS

81 Cust River on Swannonoa Rd M35 715652 1 1 9 3 PL

82 Dockey's Stream at McIntoshs Rd M35 690673 1 1 1 1 PS

83 Cust River on Oxford-Rangiora Rd M35 661661 11 2 12 2 PM

84 Hunters Stream on Boundary Rd M35 669652 9 1 3 2 PT

85 Hunters Stream on Springbank Rd M35 650638 9 1 4 2 PS

86 Drainage ditch on Boundary Rd M35 628639 5 1 4 2 PT

87 Drain on Gartery's Rd M35 632645 1 1 2 1 PT

88 Trib into Cust River on Patersons Rd M35 629675 1 1 1 1 PS

89 Cust River on Patersons Rd M35 628670 6 1 8 2 PM

90 Cust River at Swamp Rd, near Cust M35 603664 8 2 8 2 PM

91 Cust River on Tippings Rd M35 576671 12 2 12 2 PS

92 Cust River on Bennetts Rd M35 535678 8 2 8 2 PM

93 Cust River at Carleton Rd M35 503692 9 1 1 1 PM

94 Eyre River at Steffens Rd M35 507637 1 1 9 3 PL

37 Okuti River - top bridge on Okuti Valley Rd N36 977137 12 2 12 2 VS

38 Tributary into Okuti River N36 973136 12 2 12 2 VS

39 Tributary into Mid-Okuti River N36 958134 6 1 12 2 VM

95 Okuti River at Ushers Rd N36 945134 6 1 12 2 PM

96 Te Oka Stream near beach N37 927066 1 1 8 2 UM

97 Stream at Tumbledown Bay N37 917063 3 2 12 2 UT

98 Okuti River at Kinloch Rd N36 935134 11 2 12 2 PS

99 Hukahuka Turoa Stream at waterlevel recorder N36 937175 8 2 12 2 PM

100 Hukahuka Turoa Stream at Montgomeries Rd N36 933188 3 2 12 2 US

101 Hikuika Stream at end of Whites Rd N36 968208 12 2 12 2 VS

102 Hikuika Stream on Puaha Rd N36 965181 9 1 12 2 PS

103 Opuahou Stream on Puaha Rd N36 967184 10 1 12 2 PS

104 Pigeon Bay Stream at Kukupu N36 018208 11 2 12 2 VS

105 Pigeon Bay Stream at Pigeon Bay Rd N36 018212 6 1 8 2 US



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

106 Pigeon Bay Stream at Wilsons Rd N36 018225 2 1 12 2 US

107 Totara Stream, trib into Pigeon Bay Stream N36 016233 9 1 8 2 PS

108 Holmes Stream at Port-Levy Pigeon Bay Rd N36 002258 5 1 12 2 UM

109 Te Kawa Stream at Port Levy N36 952273 2 1 12 2 PM

110 Te Kawa Stream on Port-Levy Little River Rd N36 940262 5 1 8 2 US

111 Tributary into Te Kawa Stream N36 923235 12 2 12 2 VS

JQ1 Heathcote River at Pioneer Park M36 787375 4 2 12 2 PS

JQ2 Heathcote River at Warren Crescent, Curtletts Reserve M36 761390 12 2 12 2 PS

JQ3 Nottingham Stream at Nichols Rd M36 747363 8 2 7 2 PS

JQ4 Small tributary into Halswell River M36 779315 12 2 4 2 VT

JQ5 Tributary into Halswell River, Early Valley Rd M36 768322 12 2 3 2 US

JQ6 Lower tributary into Halswell River, Early Valley Rd M36 755319 5 1 7 2 PT

JQ7 Halswell River at Osterholts Rd M36 747312 1 1 7 2 PS

JQ8 Halswell River at Tai Tapu bridge M36 735274 11 2 12 2 PM

JQ9 Cam River at SH 1 bridge M35 824614 10 1 8 2 PM

JQ10 Cam River off Tuahiwi Rd M35 814623 3 2 11 2 PM

JQ11 Cam River at Bramlegs Rd M35 805625 10 1 11 2 PM

JQ12 Cam River at Waikorara Rd M35 801650 10 1 3 2 PS

JQ13 North Brook River at Boys Rd M35 788657 5 1 8 2 PS

JQ14 Ashley River at Cones Rd Bridge M35 764696 1 1 9 3 PL

JQ15 Makerikeri River on Dixon's Rd bridge M34 735709 5 1 9 3 PL

JQ16 Makerikeri River on Station Rd M34 708760 10 1 11 2 PS

JQ17 Marerikeru River at Ford M34 721785 6 1 10 3 US

JQ18 Trib into Marerikeru River on Terrace Rd M34 703776 5 1 4 2 PS

JQ19 Grey River at Mt Grey Rd ford M34 685817 3 2 9 3 PM

JQ20 West branch of Grey River at Mt Grey ford M34 684817 3 2 10 3 UM

JQ21 Grey River at White Rock Rd M34 668785 6 1 9 3 PL

JQ23 Tracey River, tributary of Grey M34 668796 12 2 3 2 PS

JQ24 Keretu River at Loburn-Whiterock Rd bridge M34 651809 3 2 10 3 PM

JQ25 West branch of Karetu River on Taffes Glen Rd M34 646817 9 1 8 2 PT

JQ26 Kowhai Stream at Taffes Glen Rd M34 626835 6 1 2 1 PS

JQ27 Fox Creek on Taffes Glen Rd M34 605845 8 2 11 2 US

JQ28 Okuku River at Whiterock Downs station M34 607845 1 1 9 3 UL

E1 Heathcote River @ Pioneer Park M36 787375 4 2 11 2 PS

E2 Halswell River at Osterholts Rd M36 747312 2 1 8 2 PS

E3 Okana Stream M36 847174 1 1 3 2 PT

E4 Saltwater at SH1 (tidal) J39 700417 2 1 11 2 PL



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

E5 Saltwater Creek @ Fairview Rd J39 684428 10 1 12 2 PL

E6 Otipua @ Brushfield Rd J38 677643 12 2 4 2 VS

E7 Saltwater @ Brockley Rd J39 625429 11 2 4 2 VT

E8 Waikakahi @ Te Maihoroa Rd J41 595861 10 1 7 2 PS

E9 Waikakahi @ Glenavy Tawai Rd (Upstream only) J41 579868 4 2 8 2 PS

E10 Waikakahi @ Cock and Hen J40 491909 10 1 3 2 PT

E11 Awakino @ ford on Awakino Rd I40 059048 6 1 5 3 UM

E12 Little Awakino River @ Ford Awakino Rd I40 038057 1 1 4 2 UT

E13 Awakino East branch I40 040018 1 1 9 3 VM

E14 West Branch of Awakino I40 040019 1 1 9 3 VS

E15 Spring Creek SH8 (non willow stretch) H39 726478 3 2 10 3 PT

E16 Twizel SH8 H38 793573 4 2 10 3 PL

E17 Frasers Stream d/s ford H38 751605 3 2 10 3 PM

E18 Dry Stream @ Canal Rd (downstream) H38 762618 1 1 9 3 PS

E19 Bullock Creek Culvert Upstream I38 168769 4 2 9 3 PT

E20 Bains Crossing @ SH8 downstream of culvert J38 346782 1 1 2 1 PS

E21 LI @ Lincoln M36 685297 12 2 12 2 PT

E22 LII @ McDonalds Rd bridge (upstream) M36 681257 2 1 6 3 PM

E23 LII @ McDonalds Rd bridge (downstream) M36 681257 1 1 11 2 PM

E24 Selwyn @ Coes Ford M36 625232 4 2 11 2 PL

E25 Irwell @ ChCh Leeston Rd (upstream) M36 575205 11 2 3 2 PT

E26 Hanmer Rd Drain M36 567189 5 1 1 1 PS

E27 Harts @ Leeston and Lake Rd M36 558118 7 2 12 2 PM

E28 Drain on Heslerton Rd (downstream) L36 400228 10 1 1 1 PT

E29 Drain on Heslerton Rd (upstream) L36 400228 10 1 3 2 PT

E30 Drain on Herleston Rd (downstm only) L36 441203 9 1 1 1 PT

E31 Drain on Herleston Rd (upstm only) L36 441203 1 1 1 1 PT

E32 Hororata on Derretts Rd (upstm only) L36 281376 3 2 11 2 PM

E33 Drain on Cordy's Rd (downstm only) L36 244399 1 1 2 1 PT

E34 Selwyn @ Bealey Rd (downstm only) L36 314390 1 1 9 3 PL

E35 Hawkins @ SH72 (upstm only) L35 317542 10 1 2 1 PM

E36 Selwyn @ SH72 (upstm only) L35 230467 8 2 11 2 PL

E37 Acheron @ Lake Coleridge Rd K35 955555 3 2 6 3 PS

E38 Trib of Camp Gully Stm on Coleridge Rd (upstm only) K35 020455 1 1 1 1 PL

E39 Camp Gully Stm @ SH72 K35 027414 6 1 9 3 PM

E40 Unnamed trib of Ashburton River @ SH72 (upstm only) K36 917326 1 1 2 1 PT

E41 Ashburton @ SH72 North Branch (downstm only) K36 915324 3 2 9 3 PL



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

E42 Staverly Stm @ SH72 (downstm only) K36 853277 10 1 2 1 PT

E43 Bowyers Stm @ SH72 (downstm only) K36 846236 8 2 10 3 PM

E44 Drain/Race on Tramway Rd K36 894188 1 1 1 1 PL

E45 Drain on Thompson Track Rd K36 075174 5 1 1 1 PT

E46 Harding Creek @ Lyndhurst Culvert K36 016135 6 1 7 2 PT

E47 Stm on Thompson Track Rd (downstm only) K36 939183 10 1 3 2 PT

E48 Stm on Thompson Track Rd (upstm only) K36 939183 1 1 3 2 PT

E49 Ashburton River South Branch at Buicks Bridge Hakatere Heron Rd (upstream only). J36 615344 6 1 9 3 PL

E50 Puddle Hill Creek at Hakatere Heron Rd(downstream only) J36 623319 10 1 6 3 PS

E51 Potts River at Hakatere Potts Rd J36 446348 1 1 9 3 PL

E52 Whiskey Creek at Hakatere Potts Rd (upstream only) J36 493320 3 2 6 3 PS

E53 Lambies Stream at Hakatere Potts Rd J36 576308 9 1 6 3 PS

E54 Woolshed Creek at Ashburton Gorge Rd (downstream only) K36 762240 11 2 2 1 PM

E55 Woolshed Creek at Ashburton Gorge Rd (upstream only) K36 762240 2 1 5 3 PL

E56 Drain on Montatts Rd (dry) K36 709141 11 2 1 1 PT

E57 Hinds at SH72 below bridge (downstream only) K37 836097 4 2 9 3 PM

E58 Drain on Mill Rd ( upstream only) K37 019043 5 1 1 1 PT

E59 Drain on Rules Rd (downstream only) L37 276041 1 1 1 1 PT

E60 Wakanui on Corbetts RD (downstream only) L37 189883 5 1 2 1 PT

E61 Hinds at SH7 K37 961982 8 2 10 3 PL

E62 Drain on Surveyors Rd (upstream only) K37 017849 5 1 11 2 PT

E63 Drain on Surveyors Rd (downstream only) K37 017849 9 1 2 1 PT

E64 Drain on Boltons Rd (downstream only) K37 822953 1 1 2 1 PT

E65 Coopers Creek at SH72 (downstream only) K37 719865 10 1 7 2 PS

E66 Coopers Creek at SH72 (upstream only) K37 719865 3 2 10 3 PS

E67 Black Creek SH 79 (upstream only) J38 659749 10 1 2 1 PT

E68 Black Creek SH79 (downstream only) J38 659749 5 1 1 1 PT

E69 Trib of Kakahu on Pletcher Rd J38 618698 5 1 3 2 PT

E70 Rangitira at Seven Sisters Rd above bridge (upstream only) J38 662667 12 2 3 2 PT

E71 Raupo Creek on Walker Rd J38 664625 9 1 2 1 PS

E72 Kohika at Twinstock Rd (partly dried up) J39 579253 9 1 2 1 PT

E73 Sir Charles Creek at Lindsays Pass Road,below road J40 633033 11 2 7 2 PT

E74 Trib of Waihao at Gum Tree Plat Rd J40 528995 5 1 2 1 PT

E75 Papaka Stream at Washdyke Rd (downstream only and almost stagnant) J39 687487 5 1 9 3 PS

E76 Unnamed trib of Waihao River at Mt Harris Rd(partly dried up) J40 478995 12 2 2 1 PT



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

E77 Trib of South Branch Waihao River at Waihaorunga Rd J40 373049 9 1 6 3 PT

E78 Sth branch Waihao River at Kaiwarua Rd (downstream only) J40 350148 5 1 1 1 VS

E79 Stony Creek on Stony Creek Rd ( downstreamonly) J40 437047 10 1 2 1 PT

E80 Waikokopara at SH82 (downstream of bridge) J40 534036 9 1 2 1 PS

E81 Waimate Creek at Mill Rd J40 532070 6 1 5 3 PS

E82 Hook Stream at Waimate Hunter Rd (downstream only) J40 543113 10 1 3 2 PT

E83 Hook River at Waimate Hunter Rd J40 531152 7 2 9 3 PM

E84 Makikihi on Pakihi Rd J40 535192 10 1 5 3 PS

E85 Teschemaker at Back line Rd (dried up. Below bridge only) J39 505237 11 2 5 3 PL

E86 Otaio River at School Crossing RD (partly dry, almost all) J39 490312 7 2 9 3 PT

E87 Unnamed trib of Paeora River at Paeora River Rd (dry) J39 598370 1 1 9 3 PT

E88 Trib of Washdyke on Kings Rd (almost dry) J39 669484 11 2 1 1 PT

E89 Oaklands Stream on Kelands Hill Rd (partly dry, stagnant) J39 670481 9 1 1 1 PT

E90 Rosewill Stream at Rosewill Valley Rd (partlydry, stagnant) J39 665499 9 1 3 2 PT

E91 Papaka at Levels Store Rd (dried up) J38 666515 1 1 1 1 PT

E92 Papaka at Connells Rd J38 626543 11 2 2 1 PT

E93 Trib of Papaka stream on Marshall Rd (downstream only, dry) J38 581536 1 1 1 1 PT

E94 Rosewill at Bassett Rd (partly dried up) J38 620514 10 1 3 2 PT

E95 Oakwood at Brockley Rd(downstream only, stagnant, partly dry) J39 613482 5 1 1 1 PT

E96 Taiko at Paeora Ford Rd (partly dry) J39 549480 11 2 2 1 PT

E97 Trib of Pig Hunting Creek at Briens Rd (dry) J39 585437 12 2 2 1 PT

E98 Trib of Pig Hunting Creek at Holme Station Rd (upstream only) J39 593410 1 1 9 3 PT

E99 Pig Hunting Creek at George Ward Rd (upstream only) J39 627404 11 2 3 2 PT

E100 Pig Hunting Creek at SH8 J39 688368 9 1 2 1 PS

E101 Unnamed stream (semi dry, stagnant) J39 657409 9 1 1 1 PT

E102 Otipua Creek North Branch at Claremont Rd Reserve (only just moving) J39 660451 12 2 11 2 PT

E103 Sutherlands Creek SH8 (dry) J38 544570 1 1 5 3 PT

E104 Burnetts stream at Cannington Rd (partly dried up, downstream only) J38 460524 12 2 3 2 PT

E105 Unnamed trib of Tengawai at Cricklewood Rd (upstream only) J38 345669 1 1 1 1 PT

E106 Unnamed trib of Tengawai River at Cricklewood Rd (downstream only) J38 347670 12 2 3 2 PT

E107 Coal Stream SH8 (downstream only) J38 371724 11 2 2 1 PS

E108 Station Creek on Plantation Rd J40 366927 7 2 5 3 PS

E109 Deep Creek at Plantation Rd J40 381952 11 2 2 1 PS



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

E110 Ross Stream at Clayton bridge on Lochaber Rd J40 404983 10 1 2 1 PT

E111 Three Springs Creek at Three Springs Rd. I38 298799 2 1 3 2 PS

E112 Halls Stream at Nixons Rd (dry) J38 339768 12 2 2 1 PS

E113 Raincliff Stream at Elliots Bridge (very slow moving) J38 456793 7 2 2 1 PS

K1 Unnamed AS1 N33 183108 1 1 7 2 VM

K2 Hurunui R @ SH1 bridge N33 179121 4 2 10 3 VL

K3 Unnamed AS2 N33 286168 9 1 5 3 PT

K4 Leader River @SH1 O32 348354 8 2 9 3 PL

K5 Conway River O32 372428 3 2 9 3 PL

K6 Waingaro Stream O32 353430 8 2 8 2 VS

K7 Limestone Stream O32 475442 5 1 4 2 PM

K8 Conway River lower O32 463442 4 2 10 3 PL

K9 Ploughman Creek at Conway Plat Rd O32 462389 9 1 4 2 VT

K10 left branch, Big Bush Gulley Stm O32 456381 4 2 10 3 VS

K11 right branch, Big Bush Gulley Stm O32 456380 4 2 10 3 UT

K12 below confluence of left and right branches O32 458379 4 2 12 2 US

K13 Silvery Crk @ Conway Plat Rd O32 473410 1 1 6 3 PS

K14 Hundalee Stm @ SH1 O32 452476 12 2 12 2 VS

K15 Limestone Stm @SH1 O32 461486 8 2 12 2 VS

K16 Okarahia Stm O32 464510 4 2 11 2 VL

K17 T Moto ?moki Stm @ Laidlaws Rd O32 498540 11 2 8 2 US

K18 stm before Paraititahi tunnel O31 576620 12 2 12 2 VT

K19 Kahutara River@ SH1 O31 584635 6 1 10 3 PL

K20 Flags Crk @SH1 P30 5197 9 1 4 2 PT

K21 Wordside Crk @ SH1 P30 999191 10 1 5 3 VL

K22 Kekerengu R @ lower rd bridge P30 927113 8 2 10 3 PL

K23 trib into Kekerengu R P30 926112 1 1 4 2 VT

K24 Valhalla Stm P30 919104 1 1 9 3 VS

K25 Shingle Fan Bridge No.2 @ SH1 P30 865989 1 1 9 3 PM

K26 McLean Stm above pine plantation P30 78723 1 1 6 3 US

K27 lower site of McLean Stm P30 78920 3 2 9 3 UL

K28 Ohau Stm @ SH1 P31 783844 12 2 12 2 VS

K29 Kowhai River @ ford O31 610686 3 2 9 3 PL

K30 Kahutara River@ dairy farm O31 545665 2 1 9 3 PM

K31 Cribb Crk @ SH70 O31 521691 3 2 9 3 PL

K32 Katutara River @ SH70 O31 470690 3 2 9 3 UL

K33 Trib into Katutara R O31 465696 4 2 9 3 US

K34 Great Burn @ Scott's Rd bridge O31 474669 12 2 8 2 US



No. Grid Ref.

RMC-C Cell

RMC-C Group

RMC-P Cell

RMC-P Group

Geomorphic Class

K35 Green Burn @ SH70 O31 431680 3 2 4 2 UM

K36 Charwell River @ SH70 bridge O31 399651 1 1 9 3 UL

K37 Stag Stm O32 391565 7 2 12 2 US

K38 Conway R @ Cloudy Range Rd bridge O31 325622 2 1 9 3 VL

K39 Conway R @ SH70 bridge O31 325607 2 1 9 3 VL

K40 Mason River N32 241561 12 2 11 2 US

K41 Mason River @ road bridge N32 236558 3 2 9 3 UL

K42 Wandel R @ SH70 N32 182468 1 1 9 3 PL

K43 Lottery R @SH70 N32 170442 1 1 9 3 PM

K44 Dog brook N32 126405 5 1 4 2 PM

K45 Blind Stm on Leslie Hills Rd N32 47414 1 1 3 2 PM

K46 Stm on Jack's Pass N32 958594 8 2 12 2 UT

K47 Peter's Valley Stm N31 933627 6 1 5 3 UM

K48 Timm's Stm @ bridge N31 925717 2 1 9 3 US

K49 Hornble Stm @ Tophouse Rd N31 931665 3 2 2 1 PS

K50 Pass Stm N31 9602 1 1 9 3 UM

K51 Dog Stm on Jollies Rd N32 966543 12 2 12 2 US

K52 Hamner River off Hossack Rd N32 998489 6 1 9 3 PM

K53 Brown's Stm N32 907380 2 1 10 3 VS

K54 Countess Stm @ SH7 N32 984327 1 1 1 1 PS

K55 Stanton Stm by Leader-Waiau Rd N32 209410 1 1 1 1 US

K56 Stanton Stm by road bridge N32 216419 12 2 4 2 US

K57 Stanton Stm near sheep yard N32 234427 1 1 3 2 US

K58 Leader River @ Mendip Hills Stm O32 326404 3 2 10 3 PL

K59 Motanau R @ Buchanans Rd N34 158959 12 2 8 2 UL

K60 Motanau R in gorge N34 132970 8 2 11 2 VM

K61 Motanau R @ Sudbury N34 106980 12 2 8 2 US

K62 Cave Crk @ Motanau Beach Rd N33 105008 2 1 1 1 US


8. Appendix 2:

Representative cross-section sketches of geomorphic riparian management classes.

riparian management classification for canterbury streams · riparian management classification for...

Documents